YARN TRAVELLING INFORMATION ACQUIRING DEVICE, YARN PROCESSING DEVICE,
AND YARN TRAVELLING INFORMATION ACQUIRING METHOD
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a yarn travelling information
acquiring device for detecting a state of a travelling yarn and a
yarn processing device equipped with the yarn travelling information
acquiring device.
10 2. Description of the Related Art
In a yarn winding device for winding a yarn, information
relating to a travelling state of the yarn is sometimes necessary
to control the winding of the yarn. Such a yarn winding device thus
includes a yarn travelling information acquiring device for acquiring
15 information relating to the travelling state of the yarn (yarn
travelling information). The travelling state may include, for
example, a travelling speed of the yarn.
Such yarn travelling information acquiring device is described,
for example, in Japanese Unexamined Patent Publication No.
20 2012-051672. The yarn travelling information acquiring device
disclosed in the document includes two yarn thickness unevenness
sensors for detecting yarn thickness unevenness, and the signals
detected by the two yarn thickness unevenness sensors are sampled
at a predetermined sampling period (cycle). The yarn travelling
25 information acquiring devices detect the yarn travelling speed by
comparing the two obtained yarn thickness unevenness signals.
BRIEF SUMMARY OF THE INVENTION
For example, since the yarn winding machine winds a yarn having
30 a very long length, the error in the length of the wound yarn becomes
large even if the error in the detection of the yarn travelling speed
by the yarn travelling information acquiring device is very small.
The yarn travelling information acquiring device capable of more
accurately acquiring the travelling state of the yarn is desired.
35 However, if the resolution is simply increased, the processing amount
increases and this is not preferable for the yarn travelling
information acquiring device.
It is an object of the present invention to provide a yarn
travelling information acquiring device capable of accurately
5 acquiring the travelling state of the yarn while reducing the
processing load.
A yarn travelling information acquiring device includes a first
detecting section adapted to detect a thickness unevenness of a
travelling yarn and to output first yarn thickness unevenness signals;
10 a second detecting section arranged upstream in a yarn travelling
direction at a distance from the first detecting section, and adapted
to detect the thickness unevenness of the yarn and to output second
yarn thickness unevenness signals; a similarity degree evaluating
section adapted to use a first imaginary frame and a second imaginary
15 frame to select a plurality of reference positions of the first
imaginary frame on a time axis of the second imaginary frame, and
to evaluate a plurality of similarity degrees of the first yarn
thickness unevenness signals and the second yarn thickness unevenness
signals, the first imaginary frame being formed in accordance with
20 the first yarn thickness unevenness signals extracted from the first
yarn thickness unevenness signals acquired within a first time range,
and the second imaginary frame being formed in accordance with the
second yarn thickness unevenness signals extracted from the second
yarn thickness unevenness signals acquired within a second time range
25 that is longer than the first time range; and a travelling information
acquiring section adapted to calculate a time delay between the first
yarn thickness unevenness signals forming the first imaginary frame
and the second yarn thickness unevenness signals forming the second
imaginary frame in accordance with the similarity degrees, and to
30 acquire travelling information of the yarn in accordance with the
distance and the time delay.
A yarn processing device includes the yarn travelling
information acquiring device; a yarn processing section adapted to
perform a processing on the yarn; and a control section adapted to
35 control the processing of the yarn processing section in accordance
with the travelling information of the yarn acquired by the yarn
travelling information acquiring device.
A yarn travelling information acquiring method includes the
following steps: detecting a thickness unevenness of a travelling
5 yarn by a first detecting section and generating first yarn thickness
unevenness signals; detecting the thickness unevenness of the yarn
by a second detecting section located upstream in a yarn travelling
direction at a distance from the first detecting section, and
generating second yarn thickness unevenness signals; using a first
10 imaginary frame and a second imaginary frame, selecting a plurality
of reference positions of the first imaginary frame on a time axis
of the second imaginary frame, and evaluating a plurality of
similarity degrees of the first yarn thickness unevenness signals
and the second yarn thickness unevenness signals, forming the first
15 imaginary frame in accordance with the first yarn thickness unevenness
signals extracted from the first yarn thickness unevenness signals
acquired within a first time range, and forming the second imaginary
frame in accordance with the second yarn thickness unevenness signals
extracted from the second yarn thickness unevenness signals acquired
20 within a second time range that is longer than the first time range;
and calculating a time delay between the first yarn thickness
unevenness signals forming the first imaginary frame and the second
yarn thickness unevenness signals forming the second imaginary frame
in accordance with the similarity degrees, and generating travelling
25 information of the yarn in accordance with the distance and the time
delay.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a side view of a winder unit;
30 FIG. 2 is a front view of the winder unit;
FIG. 3 is a block diagram illustrating a configuration of a
clearer;
FIGS. 4 are graphs illustrating data series accumulated in a
ring buffer;
35 FIG. 5 is a view for describing a calculation frame;
4
FIG. 6 is a view describing bias component removal and
normalization;
FIG. 7 is a graph illustrating a case in which a plurality of
peaks of similarity degrees exist;
5 FIG. 8 is a graph in which only maximum points of the similarity
degrees are extracted;
FIG. 9 is a flowchart of yarn travelling speed acquiring
processing;
FIG. 10 is a flowchart of an adoption determination processing;
10 FIG. 11 is a view for describing weighting by a weighting curve;
FIG. 12A and FIG. 12B are graphs illustrating examples in which
the weighting is performed on the similarity degree;
FIG. 13 is a view illustrating a state in which a downstream
frame is moved;
15 FIG. 14A to FIG. 14D are views illustrating a calculation
procedure for calculating a history weighting curve; and
FIG. 15A to FIG. 150 are views illustrating a procedure for
calculating a sampling speed weighting curve.
20 DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
Preferred embodiments of the present invention will be
described below with reference to the drawings. FIG. 1 is a side view
of a winder unit (yarn processing device) 10 arranged in an automatic
winder according to one embodiment of the present invention. FIG.
25 2 is a front view illustrating a schematic configuration of the winder
unit 10.
The winder unit 10 illustrated in FIG. 1 and FIG. 2 unwinds a
spun yarn 20 from a yarn supplying bobbin 21 and winds the spun yarn
20 around a winding bobbin 22 while traversing the spun yarn 20 to
30 form a package 30 of a prescribed length and a prescribed form. The
automatic winder of the present embodiment includes a plurality of
winder units 10 arranged in a line, and a main control device (not
illustrated) arranged at one end in a direction in which the winder
units 10 are arranged.
35 Each of the winder units 10 includes a unit frame 11 (FIG. 1)
5
arranged on a left or right side in front view, and a winding unit
main body 16 arranged at a side of the unit frame 11. The winding
unit main body (yarn processing section) 16 includes a winding section
31. The winding unit main body 16 includes a magazine-type supplying
5 device 60, and a supplying bobbin holding section 71.
As illustrated in FIG. 1, the magazine-type supplying device
60 includes a magazine holding section 61 that extends diagonally
from a lower part of the winder unit 10 towards an upward direction
and towards the front, and a bobbin accommodating device 62 that is
10 attached to a distal end of the magazine holding section 61. The
bobbin accommodating device 62 includes a magazine can 63. A
plurality of accommodation holes to which the yarn supplying bobbins
70 can be respectively set are formed in the magazine can 63. The
magazine can 63 can be intermittently driven, rotated, and fed by
15 a motor (not illustrated) , and the supply bobbins 70 can be dropped
one at a time to a bobbin supply path (not illustrated) of the magazine
holding section 61 by the intermittent drive and a control valve (not
illustrated) of the magazine can 63. The supply bobbin 70 is then
guided to the supplying bobbin holding section 71.
20 In place of the magazine-type supplying device 60 illustrated
in FIG. 1, a transport conveyor (not illustrated) arranged at a lower
part of the automatic winder may be used to supply the yarn supplying
bobbin 21 from a yarn supplying bobbin supplying section (not
illustrated) to the supplying bobbin holding section 71 of each winder
25 unit 10.
The winding section 31 winds the yarn 20, which has been unwound
from the yarn supplying bobbin 21, around the winding bobbin 22 to
form the package 30. Specifically, the winding section 31 includes
a cradle 23 configured capable of holding the winding bobbin 22, and
30 a winding drum 24 for traversing the spun yarn 20 and driving the
winding bobbin 22. The cradle 23 can be swung in a direction of
approaching or separating with respect to the winding drum 24. The
package 30 is thus brought into contact with or separated from the
winding drum 24. As illustrated in FIG. 2, a spiral-shaped traverse
35 groove 27 is formed on an outer circumferential surface of the winding
6
drum 24, and the spun yarn 20 is traversed by the traverse groove
27.
The winding bobbin 22 is rotated by driving and rotating the
winding drum 24 arranged facing the winding bobbin 22. The spun yarn
5 20 is wound around the rotating winding bobbin 22 while being traversed
by the traverse groove 27. As illustrated in FIG. 2, the winding drum
24 is coupled to an output shaft of a drum drive motor 53 . An operation
of the drum drive motor 53 is controlled by a motor control section
54. The motor control section 54 performs control to operate and stop
10 the drum drive motor 53 in response to a control signal from a unit
control section (control section) 50.
A rotation sensor 42 is attached to the winding drum 24. The
rotation sensor 42 is electrically connected to an analyzer 52 or
the like of a clearer 15, to be described later. The rotation sensor
15 42 is configured as a rotary encoder, for example, and transmits a
pulse signal to the analyzer 52 every time the winding drum 24 rotates
by a prescribed angle. The pulse signal output by the rotation sensor
42 is referred to as a rotation pulse signal.
The winding unit main body 16 has a configuration in which an
20 unwinding assisting device 12, a tension applying device 13, a yarn
joining device 14, a clearer head 49 of the clearer (yarn travelling
information acquiring device) 15 are arranged in this order from the
yarn supplying bobbin 21 side on a yarn travelling path between the
yarn supplying bobbin 21 and the winding drum 24.
25 The unwinding assisting device 12 assists the unwinding of the
spun yarn 20 from the yarn supplying bobbin 21 by lowering a regulating
member 4 0 covering a core tube of the yarn supplying bobbin 21
accompanying unwinding of the spun yarn 20 from the yarn supplying
bobbin 21. The regulating member 40 makes contact with a balloon that
30 is formed at an upper part of the yarn supplying bobbin 21 when the
spun yarn 20 unwound from the yarn supplying bobbin 21 is swung around,
thus applying an appropriate tension to the balloon and assisting
unwinding of the spun yarn 20.
The tension applying device 13 applies a prescribed tension to
35 the travelling spun yarn 20. The tension applying device 13 applies
7
constant tension to the spun yarn 20, thereby improving quality of
the package 30.
The clearer 15 detects defects by detecting thickness
unevenness of the spun yarn 20 with an appropriate sensor.
5 Specifically, the clearer 15 includes the clearer head 49 and the
analyzer 52 (FIG. 2). Two yarn unevenness detecting sensors 43 and
44 are arranged in the clearer head 49. Yarn defects such as slub
can be detected by processing signals from the yarn unevenness
detecting sensors 43 and 44 by the analyzer 52. A cutter (not
10 illustrated) for immediately cutting the spun yarn 20 when the clearer
15 detects a yarn defect is arranged in proximity to the clearer head
49. ^
The clearer 15 may also function as a yarn travelling
information acquiring device for acquiring travelling information
15 of the spun yarn 20. The travelling information of the spun yarn 20
is information indicating a state of the travelling spun yarn 20.
The configuration for acquiring the travelling information of the
spun yarn 20 by the clearer 15 will be described later.
After the clearer 15 detects a yarn defect and cuts the spun
20 yarn 20, or after a yarn breakage during unwinding of the spun yarn
20 from the yarn supplying bobbin 21, and the like, the yarn joining
device 14 joins a lower yarn from the yarn supplying bobbin 21 and
an upper yarn from the package 30. The yarn joining device 14 may
be a mechanical-type or a type that uses fluid such as compressed
25 air.
A lower yarn guiding pipe 25 adapted to catch and guide the lower
yarn from the yarn supplying bobbin 21 and an upper yarn guiding pipe
26 adapted to catch and guide the upper yarn from the yarn supplying
bobbin 30 are arranged on the lower side and the upper side,
30 respectively, of the yarn joining device 14. A suction port 32 is
formed at a tip end of the lower yarn guiding pipe 25. A suction mouth
34 is arranged at a tip end of the upper yarn guiding pipe 26. The
lower yarn guiding pipe 25 and the upper yarn guiding pipe 26 are
respectively connected to an appropriate negative pressure source
35 (not illustrated) , and a suction flow can be generated at the suction
port 32 and the suction mouth 34.
At the time of yarn cut or yarn breakage, the suction port 32
of the lower yarn guiding pipe 25 catches the lower yarn at a position
illustrated in FIG. 1 and FIG. 2, and is thereafter swung upward around
5 a shaft 33 to guide the lower yarn to the yarn joining device 14,
Almost at the same time, the upper yarn guiding pipe 26 is swung upward
around a shaft 35 from the illustrated position to catch, with the
suction mouth 34, the upper yarn unwound from the package 30.
Subsequently, the upper yarn guiding pipe 26 is swung downward around
10 the shaft 35 to guide the upper yarn to the yarn joining device 14.
The yarn joining operation of the lower yarn and the upper yarn is
then carried out by the yarn joining device 14,
Next, the clearer 15 will be described in detail with reference
to FIG. 3.
15 As illustrated in FIG. 3, the clearer head 49 includes a first
yarn unevenness detecting sensor (first detecting section) 43, a
second yarn unevenness detecting sensor (second detecting section)
44 and two A/D converters 45 and 46, The analyzer 52 is configured
by hardware such as a Central Processing Unit (CPU) 47, a Random Access
20 Memory (RAM) 48, and a Read Only Memory (ROM) (not illustrated), and
software such as program stored in the ROM, The hardware and the
software cooperate to cause the CPU 47 to function as a similarity
degree evaluating section 65, a weighting processing section 66, a
travelling information acquiring section 67, a yarn quality measuring
25 section 68, a sampling speed acquiring section 72, a measuring section
73, and the like. The pulse signals from the rotation sensor 42 are
input to the analyzer 52.
The first yarn unevenness detecting sensor 43 and the second
yarn unevenness detecting sensor 44 are arranged at a prescribed
30 interval in a yarn travelling direction. The first yarn unevenness
detecting sensor 43 is arranged downstream and the second yarn
unevenness detecting sensor 44 is arranged upstream. In the present
embodiment, the yarn unevenness detecting sensors 43 and 44 are
adapted to detect thickness unevenness of the spun yarn 20.
35 Specifically, the yarn unevenness detecting sensors 43 and 44 are
configured as optical sensors. Light Emitting Diodes (LED) 36 and
37 are arranged as light sources on an opposite side of the yarn
unevenness detecting sensors 43 and 44, respectively, with a yarn
path of the spun yarn 20 therebetween. The yarn unevenness detecting
5 sensors 43 and 44 detect a light receiving amount of the light
transmitted from the LEDs 36 and 37, respectively. Since the light
receiving amount of the yarn unevenness detecting sensors 43 and 44
changes when the thickness of the travelling spun yarn 20 changes,
the clearer 15 can detect the thickness unevenness of the spun yarn
10 20. The output signals (yarn thickness unevenness signals) from the
yarn unevenness detecting sensors 43 and 44 are Analog-to-Digital
(A/D) converted, and then output to the analyzer 52.
The CPU 47 arranged in the analyzer 52 monitors the A/D converted
yarn thickness unevenness signals and measures the quality of the
15 spun yarn 20. For example, because the thickness of the spun yarn
20 is found to be abnormal in a portion where the quality of the spun
yarn 20 has a problem, the defect of the spun yarn 20 can be detected
by detecting the abnormality in the thickness of the spun yarn 20
by the CPU 47. Since the quality of the spun yarn 20 is measured by
20 the CPU 47, the CPU 4 7 thus can be referred to be functioning as the
yarn quality measuring section 68.
The yarn supplying bobbin 21 normally has a yarn spun by a ring
spinning machine. Slight thickness unevenness may periodically
occur in such a yarn. The cause of the periodic yarn thickness
25 unevenness may include core shift of a draft roller that drafts a
sliver in the ring spinning machine. The periodic thickness
unevenness in a spinning process causes moire to be generated in a
woven cloth in the subsequent weaving process. The CPU 47 serving
as the yarn quality measuring section 68 performs a Fast Fourier
30 Transform (FFT) calculation of the yarn thickness unevenness signal
to detect the periodic thickness unevenness of the spun yarn 20. In
order to accurately perform the FFT calculation, the number of
waveform data per unit length of the spun yarn 20 is required to be
accurately made constant when sampling the yarn thickness unevenness
35 signal in the A/D converter.
10
The CPU 47 of the present embodiment acquires information
relating to a travelling state of the spun yarn 20, and changes a
sampling period (cycle) of the second A/D converter 46 according to
the travelling state. Specifically, the CPU 47 generates a pulse
5 signal each time the spun yarn 20 travels a specific length (e.g.,
1 mm) , and transmits the pulse signal to the second A/D converter
46. This pulse signal is referred to as a fixed yarn length pulse
signal. Based on this fixed yarn length pulse signal, the second A/D
converter 46 samples the analog signals from the first yarn unevenness
10 detecting sensor 43, and converts the analog signals into digital
signals. Accordingly, since the number of data per unit length of
the spun yarn 20 can be accurately maintained constant, the FFT
calculation can be accurately performed in the CPU 47 and the periodic
thickness unevenness can be reliably detected. By accurately
15 maintaining the number of data per unit length of the spun yarn 20
constant, the CPU 47 can accurately perform an evaluation of the length
of the thickness unevenness of the spun yarn 20 even with a sporadic
yarn defect without periodicity, and the detection accuracy of the
analyzer 52 can be improved. Since the fixed yarn length pulse signal
20 is information relating to the travelling state of the yarn 20, the
fixed yarn length pulse signal can be referred to as one type of yarn
travelling information. As described above, since the CPU 47
acquires the yarn travelling information, the CPU 47 may be referred
to be functioning as the travelling information acquiring section
25 67.
Next, the configuration for acquiring the fixed yarn length
pulse signal will be described.
The clearer 15 of the present embodiment includes the first A/D
converter 45 apart from the second A/D converter 46.
30 The first A/D converter 45 is an A/D converter which performs
sampling of the yarn thickness unevenness signal in order to acquire
the fixed yarn length pulse signal by the CPU 47. Specifically, the
first A/D converter 45 samples the analog signals from the two yarn
unevenness detecting sensors 43 and 44, and converts the analog
35 signals to digital signals. The obtained digital signals are input
n
into the analyzer 52. The CPU 47 arranged in the analyzer 52 functions
as the similarity degree evaluating section 65, the weighting
processing section 66, and/or the travelling information acquiring
section 67 to detect a travelling speed of the spun yarn 20 using
5 the input digital signals. The travelling speed of the spun yarn 20
is also information relating to the travelling state of the spun yarn
20, and thus can be referred to as one type of yarn travelling
information. The CPU 47 may be referred to be functioning as the
travelling information acquiring section 67.
10 If the travelling speed of the spun yarn 20 is obtained, the
travelled length of the spun yarn 20 within a prescribed period of
time can be detected in accordance with the travelling speed. The
CPU 47 generates and acquires the fixed yarn length pulse signal in
accordance with the travelling speed of the spun yarn 20, and transmits
15 the fixed yarn length pulse signal to the second A/D converter 46.
The yarn thickness unevenness signal can be sampled for every fixed
yarn length of the spun yarn 20 in the second A/D converter 46.
Next, a method for acquiring the travelling speed (yarn
travelling information) of the spun yarn 20 by the clearer 15 will
20 be described in detail.
First, in the first A/D converter 45, the analog waveforms
output from the yarn unevenness detecting sensors 43 and 44 are sampled.
A sampling frequency fsl at this time is changed as needed in
proportion to a rotation speed of the winding drum 24 . When the signal
25 waveforms of the yarn unevenness detecting sensors 43 and 44 are
sampled by the first A/D converter 45, the number of data acquired
per unit length of the spun yarn 20 thus can be maintained
substantially constant. A calculation load of the CPU 47 can be
reduced when the sampling frequency is fixed.
30 As described above, the rotation sensor 42 outputs the rotation
pulse signal every time the winding drum 24 rotates a prescribed angle.
Therefore, the number of rotation pulse signals output per unit time
is proportional to the rotation speed of the winding drum 24. The
CPU 47 of the analyzer 52 acquires the rotational information of the
35 winding drum 24 in accordance with the rotation pulse signal received
12
from the rotation sensor 42. The rotational information of the
winding drum 24 is information relating to the rotation speed of the
winding drum 24, and may be a peripheral speed of the winding drum
24, an angular speed of the winding drum 24, the number of rotation
5 pulse signals output per unit time, and the like. That is, the CPU
47 merely needs to acquire information relating to the rotation speed
of the winding drum 24 in some form in accordance with the rotation
pulse signal.
The CPU 47 obtains the sampling speed that becomes a reference
10 of the travelling speed of the spun yarn 20 according to the processing
such as multiplying a prescribed factor to the rotational information
of the winding drum 24 obtained in the above manner. The CPU 47 obtains
the sampling frequency fsl according to the obtained sampling speed,
and sets the obtained sampling frequency fsl to the first A/D converter
15 45. The sampling frequency is set such that one sampling is performed
every time the spun yarn 20 travels a prescribed unit length. The
sampling frequency fsl of the first A/D converter 45 may be changed
as needed in proportion to the rotation speed of the winding drum
24. Since the CPU 47 calculates the sampling speed based on the
20 rotational information of the winding drum 24, the CPU 47 can be
referred to as functioning as the sampling speed acquiring section
72.
The analyzer 52 has a storage region configured as a ring buffer
(a downstream ring buffer 55 and an upstream ring buffer 56) in the
25 RAM 4 8 to temporarily carry the waveform data input from the first
A/D converter 45. Specifically, data obtained by sampling the output
signal (the first yarn thickness unevenness signal) from the first
yarn unevenness detecting sensor 43 is accumulated in the downstream
ring buffer 55. The data obtained by sampling the output signal (the
30 second yarn thickness unevenness signal) from the second yarn
unevenness detecting sensor 44 is accumulated in the upstream ring
buffer 56. Although the size of the downstream ring buffer 55 and
the upstream ring buffer 56 is not particularly limited, the
downstream ring buffer 55 and the upstream ring buffer 56 can
35 respectively carry 128 data in the present embodiment.
13
FIGS. 4 are graphs illustrating data series (waveform data
series) accumulated in the ring buffers 55 and 56. In the graphs of
FIG. 4, a vertical axis represents a signal level indicated by the
waveform data, and a horizontal axis represents an index of the ring
5 buffer in which the waveform data is stored. Regarding the index of
the horizontal axis of FIG. 4, a smaller value is assigned to older
data in the ring buffer. In other words, the index in which the oldest
data is stored in the ring buffer is index [0] , and the index in which
the most recent data is stored is index [127] . Therefore, the
10 horizontal axis of FIG. 4 may also be considered as a time axis.
If the tension applied to the spun yarn 20 is constant,
stretching of the spun yarn 20 at the measurement positions of the
first yarn unevenness detecting sensor 43 and the second yarn
unevenness detecting sensor 44 is the same, and hence the same
15 waveforms are assumed to be observed in the two yarn unevenness
detecting sensors 43 and 44. Since the first yarn unevenness
detecting sensor 43 is arranged downstream of the second yarn
unevenness detecting sensor 44 in the yarn travelling direction, the
waveform of the signal output from the first yarn unevenness detecting
20 sensor 43 (the first yarn thickness unevenness signal) has a time
delay with respect to the waveform of the signal output from the second
yarn unevenness detecting sensor 44 (the second yarn thickness
unevenness signal). This delay causes the waveform data series
stored in the upstream ring buffer 56 to shift by AT in a past direction
25 (left direction of FIG. 4) in the time axis compared to the waveform
data series stored in the downstream ring buffer 55. Provided that
the time delay is AT, and a distance between the detection positions
of the two yarn unevenness detecting sensors 43 and 44 is L, a yarn
speed V can be obtained by V = L/AT ••• (1) . Therefore, the analyzer
30 52 can calculate the travelling speed of the spun yarn 20 by detecting
the time delay AT of the waveform of the first yarn thickness
unevenness signal with respect to the waveform of the second yarn
thickness unevenness signal.
In the present embodiment, the CPU 47 compares the waveform of
35 the first yarn thickness unevenness signal (waveform data series
14
accumulated in the downstream ring buffer 55) and the waveform of
the second yarn thickness unevenness signal (waveform data series
accumulated in the upstream ring buffer 56) to obtain the time delay
AT. Since a certain duration of time is assumed when the "waveform"
5 is referred, only the single waveform data sampled at a certain instant
cannot be referred to as the "waveform". When referring to comparing
the waveforms of the signals, the data series continuously acquired
within a certain time range are compared as the specific processing.
The CPU 47 compares the waveform data series (the first waveform
10 data series) continuously acquired within a prescribed first time
range of the downstream waveform data accumulated in the downstream
ring buffer 55 and the waveform data series (the second waveform data
series) continuously acquired within a prescribed second time range
of the waveform data accumulated in the upstream buffer 56.
15 As illustrated in FIG. 4, the first time range is a range from
index [64] to index [127] of the ring buffer in the present embodiment.
Therefore, the first waveform data series is configured by waveform
data of most recent 64 points among the waveform data accumulated
in the downstream ring buffer 55. The second time range is a range
20 from index [32] to index [127] of the ring buffer. Therefore, the
second waveform data series is configured by waveform data of most
recent 96 points among the waveform data accumulated in the upstream
ring buffer 56.
Next, the comparison of the waveforms carried out by the CPU
25 47 will be specifically described. The CPU 47 compares the first
waveform data series and the second waveform data series to obtain
a similarity degree of the waveform of the first yarn thickness
unevenness signal and the waveform of the second yarn thickness
unevenness signal. The similarity degree is an indication indicating
30 to what extent the two waveforms overlap (to what extent the two
waveforms are similar).
Various methods can be considered as a method for calculating
the similarity degree. In the present embodiment, the similarity
degree is calculated in the following manner. That is, the two
35 waveforms to be compared are overlapped to acquire an area between
15
the two waveform graphs (portion hatched with diagonal lines in FIG.
5) . In this case, the two waveforms do not overlap at all if the two
waveforms are completely different, and the area becomes 2. If the
two waveforms completely coincide, the area becomes 0 . With the above
5 area, the similarity degree can be calculated by the following
equation (2).
Similarity degree = 1 - (area between two waveforms) -^ 2 ••• (2)
According to the definition of the similarity degree, the two
waveforms are more different as the similarity degree is closer to
10 0, and the two waveforms are more similar as the similarity degree
is closer to 1.
Since the second waveform data series is configured by waveform
data of 96 points and the first waveform data series is configured
by waveform data of 64 points, the range of the second waveform data
15 series is set wider than the range of the first waveform data series.
In other words, the second time range is set longer than the first
time range in the present embodiment. In order to calculate the
similarity degree, the length on the time axis of the two waveforms
(length of the waveform data series) to be compared is required to
20 be equal. Thus, when comparing the first waveform data series and
the second waveform data series, the CPU 47 extracts the waveform
data series having the same length as the first wavelength data series
from the second waveform data series, and obtains the similarity
degree of the extracted waveform data series and the first waveform
25 data series.
This will be more specifically described below. The CPU 47
prepares an imaginary frame (the downstream frame and the upstream
frame) for imaginarily retrieving the waveform data series acquired
within a prescribed time range from the waveform data of the ring
30 buffers 55 and 56. The CPU 47 evaluates the overlapping (the
similarity degree) of the waveform data of the downstream frame (the
first imaginary frame) and the waveform data of the upstream frame
(the second imaginary frame).
The downstream frame (the first imaginary frame) is an imaginary
35 frame for imaginarily retrieving the waveform data series (the first
16
waveform data series) continuously acquired in the first time range
from the waveform data accumulated in the downstream ring buffer 55.
Specifically, as illustrated in FIG. 5, the downstream frame is set
to include data of most recent 64 points (the waveform data in a range
5 from index [64] to index [127] of downstream ring buffer) among the
waveform data stored in the downstream ring buffer 55. In the
following description, the index of the oldest data of the waveform
data included in the imaginary frame is expressed as a head position
of the imaginary frame. In the case of the downstream frame, the head
10 position is index [64].
The upstream frame (the second imaginary frame) is an imaginary
frame for imaginarily retrieving the waveform data series (waveform
data of 64 points) in the time range of the same length as the first
time range from the waveform data series (the second waveform data
15 series) of the upstream ring buffer 56 acquired in the second time
range. The upstream frame is set to store data of 64 continuous points
among the waveform data of the range from index [32] to index [127]
(the second time range) of the upstream ring buffer 56.
The range of the second waveform data series is wider than the
20 first waveform data series by the waveform data of 32 points towards
the past direction on the time axis. The head position of the upstream
frame thus can be set to a position shifted within 32 points in the
past direction than the head position of the downstream frame. The
difference in index between the head position of the downstream frame
25 and the head position of the upstream frame is referred to as "frame
delay amount" or simply "delay amount".
After the upstream frame is set as described above, the CPU 47
extracts the waveform data in accordance with the predetermined
extracting condition from the waveform data of 64 points in the set
30 upstream frame, and creates a new upstream frame in accordance with
the extracted waveform data. This reduces the number of waveform data
included in the upstream frame. The predetermined extracting
condition includes extracting every other waveform data included in
the upstream frame, extracting every two waveform data, and the like,
35 for example. Similarly, the CPU 47 extracts the waveform data in
17
accordance with the predetermined extracting condition from the
waveform data of 64 points in the set downstream frame, and creates
a new downstream frame in accordance with the extracted waveform data.
For example, the extracting condition is changed when the yarn
5 travelling speed (the sampling speed) exceeds a prescribed yarn
travelling speed (the prescribed sampling speed). Since the
processing time relating to the acquisition of the yarn travelling
information is substantially constant, when the sampling speed
becomes fast, the number of points to be sampled during a period of
10 time in which the yarn travelling information is being acquired
increases. In this case, the ratio at which the yarn travelling
information is acquired with respect to the number of points to be
sampled lowers. Therefore, the extracting condition is set as
follows. If the prescribed sampling speed is smaller than a first
15 threshold value, reduction is not performed. If the prescribed
sampling speed is greater than or equal to the first threshold value,
reduction is performed for every other waveform data. If the
prescribed sampling speed is greater than or equal to a second
threshold value, reduction is performed for every two waveform data.
20 After the upstream frame and the downstream frame in which the
number of waveform data is reduced are set, the CPU 47 obtains the
similarity degree, which is the indication on what extent the waveform
included in the upstream frame and the waveform included in the
downstream frame are overlapping. The output signals of the yarn
25 unevenness detecting sensors 43 and 44 conrtain bias components.
Therefore, the two waveforms may not overlap satisfactorily due to
variation in the sensitivity of the yarn unevenness detecting sensors
43 and 44. The CPU 47 thus performs a processing of bias component
removal and normalization on the waveform data of the upstream frame
30 and the downstream frame. As illustrated in FIG. 6, the bias component
removal is a processing of searching for a minimum value of the data
in the calculation frame and subtracting the minimum value from the
value of each data. Normalization is a processing of dividing the
value of each data with a total value of data in the calculation frame.
35 The area of the graph of the waveform in the calculation frame is
18
thereby normalized to 1. Since the bias component removal and the
normalization of the waveforms are carried out on the upstream
waveform data and the downstream waveform data respectively,
variation in bias component for each of the yarn unevenness detecting
5 sensors 43 and 44, as well as the difference in sensitivity for each
of the yarn unevenness detecting sensors 43 and 44 can be absorbed.
After the processing of bias component removal and
normalization are performed as described above, the CPU 47 executes
a similarity degree evaluation processing for obtaining the
10 similarity degree between the data series of the downstream frame
(the waveform of the first yarn thickness unevenness signal) and the
data series of the upstream frame (the waveform of the second yarn
thickness unevenness signal). Since the CPU 47 evaluates the
similarity degree of the two waveforms, the CPU 47 may be referred
15 to be functioning as the similarity degree evaluating section 65.
The head position (the reference position) of the upstream frame
thus can be set to a position shifted within 32 points in the past
direction than the head position of the downstream frame. The frame
delay amount (the prescribed delay amount) can be changed in the range
20 from 0 to 32. The similarity degree evaluating section 65 repeatedly
executes the similarity degree evaluation processing while changing
the frame delay amount within the above range. The similarity degree
evaluating section 65 selects a plurality of positions (the reference
positions) on the time axis of the upstream frame in the second time
25 range, and evaluates the similarity degree with respect to each of
the plurality of positions . The similarity degree evaluating section
65 thereby acquires a plurality of similarity degrees. By setting
the second time range to be longer than the first time range, the
similarity degree evaluation can be carried out for a plurality of
30 times while moving the upstream frame within the second time range.
The CPU 47 can therefore acquire a plurality of similarity degrees.
As a result of acquiring a plurality of similarity degrees for the
frame delay amount, the CPU 47 can obtain a relationship between the
delay amount and the similarity degree, as illustrated in FIG. 7,
35 for example. The CPU 47 extracts only the maximum points (similarity
19
degree maximum points) at which the similarly degree becomes maximum
from the obtained relationship between the frame delay amount and
the similarity degree, and obtains a relationship between the frame
delay amount and the similarity degree at each extracted maximum point
5 as illustrated in FIG. 8. The processing of extracting only the
maximum points is performed to suppress error from occurring when
weighting is performed on the similarity degree. The details of
weighting will be described in detail later. The maximum point is
a point which only the maximum portion of the similarity degree is
10 extracted and represents the similarity degree.
The effects of obtaining the similarity degree by reducing the
waveform data in the upstream frame and the downstream frame will
be described. The resolution of the travelling speed of the spun yarn
20 to be calculated becomes higher by sampling the output signals
15 from the yarn unevenness detecting sensors 43 and 44 according to
the sampling frequency obtained based on the rotational information
of the winding drum 24. For example, if the distance between the yarn
unevenness detecting sensors 43 and 44 is 10 mm and the sampling is
performed every time the spun yarn 20 moves 1 mm (constant length
20 of the spun yarn 20) based on the sampling speed, the resolution
becomes 1/10 times the sampling speed like 5/10, 6/10, 7/10, 8/10, . . .
At this time, if the range to obtain the similarity degree is 20 mm
(worth 20 indices) , the evaluation processing of the similarity degree
is carried out with respect to the waveform data of 20 points. If
25 the sampling is performed every time the spun yarn 20 moves 0.5 mm,
the resolution becomes 1/20 times the sampling speed, but the
evaluation processing of the similarity degree needs to be performed
with respect to the waveform data of 40 points. The resolution and
the processing time are in a trade-off relationship. The processing
30 time for the similarity degree evaluation can be shortened by reducing
the waveform data in the upstream frame and the downstream frame while
improving the resolution of the travelling speed of the spun yarn
20 by shortening the sampling interval as in the present embodiment.
When the similarity degree becomes maximum, the waveform of the
35 upstream frame and the waveform of the downstream frame are most
20
satisfactorily overlapped. In other words, the time delay AT has been
resolved between the waveform data series of the upstream ring buffer
56 and the waveform data series of the downstream ring buffer 55.
Therefore, frame delay amount when the similarity degree is maximum
5 is assumed to correspond to the time delay AT of the waveform of the
upstream ring buffer 56 and the waveform of the downstream ring buffer
55. In other words, the time delay AT of the waveforms can be
calculated based on the frame delay amount when the similarity degree
becomes maximum. The calculation of the delay AT will be specifically
10 described later.
The clearer 15 can calculate the travelling speed of the spun
yarn 20 by substituting AT obtained as described above in the equation
(1) . In this manner, the clearer 15 can obtain the travelling speed
of the spun yarn 20 based on the similarity degree of two waveforms.
15 However, the clearer 15 of the present embodiment does not calculate
the travelling speed of the spun yarn 20 using the similarity degree
as it is as described above, and instead calculates the travelling
speed of the spun yarn 20 using the weighted similarity degree, as
hereinafter described.
20 In the present embodiment, a plurality of similarity degrees
are obtained by moving the upstream frame on the time axis as described
above, but instead, the downstream frame may be moved on the time
axis. However, if the upstream frame is moved on the time axis as
in the present embodiment, the position of the downstream frame on
25 the time axis can be fixed so as to include the most recent waveform
data among the waveform data included in the downstream ring buffer
55 (in FIG. 5, the downstream frame can be fixed at a rightmost
position) . The clearer 15 thus can calculate the time delay AT using
the most recent downstream yarn thickness unevenness signal, and can
30 obtain the travelling speed of the spun yarn 20 in real time.
Since a similar state may continue in the thickness unevenness
of the travelling spun yarn 20, a similar waveform may also continue
in the signals output from the yarn unevenness detecting sensors 43
and 44. In this case, since the upstream waveform and the downstream
35 waveform overlap at a plurality of positions, as illustrated in FIG.
21
7, a plurality of frame delay amounts indicating a peak (maximum point)
having a large similarity degree exist within the movement range of
the upstream frame. If a plurality of peaks having a large similarity
degree exist within the movement range of the upstream frame, the
5 delay amount of which peak is to be used for the calculation of the
time delay AT becomes confusing and the yarn travelling speed may
be calculated using a wrong frame delay amount.
Therefore, in the present embodiment, weighting is performed
on the similarity degree to resolve the confusion of the peak of the
10 similarity degree. The yarn travelling speed acquiring processing
executed by the CPU 47 in the present embodiment will be described
below with reference to FIG. 9.
The CPU 47 executes the yarn travelling speed acquiring
processing illustrated in the flowchart of FIG. 9 every time new data
15 is sampled in the first A/D converter 45 and new waveform data is
added to the ring buffers 55 and 56. After starting the yarn
travelling speed acquiring processing, the CPU 47 reduces the waveform
in the downstream frame, and performs bias component removal and
normalization (step SlOl) .
20 The CPU 47 then performs initialization of the frame delay
amount (initialization of the position of the upstream frame) (step
S102) . In the present embodiment, the frame delay amount is
initialized to 32. The upstream frame is thus set to a position
shifted to the past corresponding to the waveform data of 32 points
25 from the downstream frame. Therefore, the upstream frame includes
the waveform data in the range from index [32] to index [95] of the
upstream ring buffer 56 (see FIG. 5).
After the position of the upstream frame is determined, the CPU
47 reduces the waveform in the upstream frame, and performs bias
30 component removal and normalization of the waveform (step S103) .
Next, the CPU 47 performs the similarity degree evaluation
processing for obtaining the similarity degree on the waveform data
in the upstream frame and the waveform data in the downstream frame
(step S104) . After calculating the similarity degree, the CPU 47
35 saves the frame delay amount of each maximum point of the similarity
22
degree (see FIG. 8) (step S105).
The CPU 47 then determines whether or not the movement range
of the upstream frame is finished (step S106). In the present
embodiment, the upstream frame is moved in the range in which the
5 frame delay amount is from 32 to 0 (i.e., range in which the upstream
frame fits within the second time range). If the movement range is
not finished (step S106: NO), the frame delay amount is reduced by
one in step S107 (upstream frame is shifted by one toward right in
FIG. 5) , and the process returns to step S103. If the movement range
10 is finished (step S106: YES), the process proceeds to step S108.
According to the above loop processing, steps S103 to S107 are
repeatedly carried out while shifting the position of the upstream
frame on the time axis. Accordingly, since the similarity degree
evaluation processing and the weighting processing are performed for
15 a plurality of times in the movement range of the upstream frame (range
in which the frame delay amount is from 0 to 32) , the CPU 47 can acquire
a plurality of similarity degrees. The data stored in step S105 is
reset every time a new yarn travelling speed acquiring processing
is started.
20 The CPU 47 then performs an adoption determination processing
(step 5108) for determining a maximum point to be adopted in the
calculation of the time delay AT of the waveform data of the upstream
ring buffer 56 and the waveform data of the downstream ring buffer
55 for a plurality of obtained maximum points.
25 The contents of the adoption determination processing will be
specifically described below with reference to the flowchart of FIG.
10. After starting the adoption determination processing, the CPU
47 performs the weighting processing on each maximum point illustrated
in FIG. 8, and performs the weighting processing for calculating the
30 weighted maximum point (step S201) . Since the weighting processing
is carried out in this manner, the CPU 47 may be referred to be
functioning as the weighting processing section 66. In the following
description, the maximum point before the weighting is sometimes
referred to as "raw maximum point", in particular if there is a need
35 to be distinguished from the weighted maximum point. The above
23
weighting processing is performed by multiplying a weighting factor
to the value of the raw maximum point. In other words, provided that
the raw maximum point for the frame delay amount c_index is Sc_index
and the weighting factor with respect to the frame delay amount c_index
5 is Wc_indexf the weighted maximum point S'c_index can be obtained with
the following equation (3) .
^ c_index " •3c_index ^ "c_index '" (-J)
The value of the weighting factor Wc_index for a certain frame
delay amount c_index is determined by a weighting curve that specifies
10 the relationship between the frame delay amount and the weighting
factor. An example of the weighting curve is illustrated in an upper
graph of FIG. 11. As illustrated in FIG. 11, the weighting curve is
set so that the weighting factor has one maximum value that becomes
a peak at the position of a prescribed frame delay amount c_index,
15 and the value of the weighting factor is set so as to gradually decrease
as it becomes distant from the peak. Therefore, by performing the
weighting on the maximum point using the weighting factor specified
with the weighting curve, the maximum point located near the peak
of the weighting curve is emphasized and the other maximum points
20 are suppressed. As a result, as illustrated in the lower graph of
FIG. 11, the emphasis (similarity degree) of the unnecessary maximum
points among the plurality of maximum points can be suppressed.
The effects of performing the weighting processing on the
maximum point will be described. For example, when performing
25 weighting using the weighting curve with respect to the raw similarity
degree of which the weighting has not been performed as illustrated
in the upper graph of FIG. 12A, the maximum point of the weighted
similarity degree, which is the similarity degree after the weighting,
may shift from the position (frame delay amount) of the maximum point
30 of the raw similarity degree, as illustrated in the lower graph of
FIG. 12A. For example, when performing weighting using the weighting
curve with respect to the raw similarity degree of which the weighting
has not been performed as illustrated in the upper graph of FIG. 12B,
the number of maximum points of the weighted similarity degree may
35 increase, as illustrated in the lower graph of FIG. 12B.
24
As illustrated in FIG. 11, by performing weighting with respect
to the maximum points in the present embodiment, occurrence of shift
in the position of the maximum point and/or increase in the number
of maximum points can be suppressed. Therefore, the weighting
5 processing can be more accurately performed in the present embodiment.
In the present embodiment, a history weighting curve and a
sampling speed weighting curve are used as the weighting curve. In
step S201, the CPU 47 performs the processing of weighting the maximum
point using the history weighting curve and the processing of
10 weighting the maximum point using the sampling speed weighting curve.
The history weighting curve is set so that the weighting factor
becomes large in proximity to the frame delay amount where the weighted
maximum point became maximum in the previous yarn travelling speed
acquiring processing. That is, since the yarn travelling speed
15 continuously changes, the yarn travelling speed acquired in the
current yarn travelling speed acquiring processing is assumed to be
not significantly different from the yarn travelling speed acquired
in the previous yarn travelling speed acquiring processing.
Therefore, the frame delay amount where the maximum point becomes
20 maximum in the current yarn travelling speed acquiring processing
has a high probability of being in proximity to the frame delay amount
where the maximum point became maximum in the previous yarn travelling
speed acquiring processing. In other words, the raw maximum point
that appears at a position away from the frame delay amount where
25 the maximum point became maximum in the previous yarn travelling speed
acquiring processing has a high possibility of being a false maximum
point that does not correspond to the yarn travelling speed. Thus,
the weighting factor is preferably reduced to a lower degree of
importance with respect to the raw maximum point located at a position
30 away from the frame delay amount where the maximum point became maximum
in the previous yarn travelling speed acquiring processing. The
history weighting curve will be described in detail later.
The sampling speed weighting curve is set such that the
weighting factor becomes large in proximity to the frame delay amount
35 corresponding to the sampling speed calculated in accordance with
25
the rotational information of the winding drum 24. In other words,
the yarn travelling speed changes around the sampling speed, and is
assumed to not significantly differ from the sampling speed.
Therefore, the time delay AT of the waveform data of the upstream
5 ring buffer 56 with respect to the waveform data of the downstream
ring buffer 55 is assumed to change in proximity to a frame delay
corresponding to the sampling speed. In other words, the raw maximum
point that appears at a position located away from the frame delay
amount corresponding to the sampling speed has a high possibility
10 of being a false maximum point that does not correspond to the yarn
travelling speed. Thus, the weighting factor is preferably reduced
to lower degree of importance with respect to the raw maximum point
located at a position away from the frame delay amount corresponding
to the sampling speed. The sampling speed weighting curve will be
15 described in detail later.
Referring back to FIG. 10, the CPU 47 calculates a plurality
of weighted maximum points using the history weighting curve and the
sampling speed weighting curve, and then extracts the weighted maximum
point having the largest similarity degree (similarity degree after
20 weighting) (step S202). The CPU 47 determines whether or not a
plurality of weighted maximum points having the largest similarity
degree are extracted (step S203) . If only one weighted maximum point
having the largest similarity degree is extracted (step S203: NO),
the CPU 47 adopts the maximum point extracted in step S202 as the
25 maximum point (target weighted similarity degree) for calculating
the time delay AT of the waveform data.
If a plurality of weighted maximum points having the largest
similarity degree are extracted (step S203: YES) , the CPU 47 compares
the similarity degrees before weighting for the extracted weighted
30 maximum points . The CPU 47 extracts the weighted maximum point having
the largest similarity degree before weighting as a result of the
comparison (step S204). The CPU 47 determines whether or not a
plurality of maximum points having the largest similarity degree
before weighting are extracted (step S205). If only one weighted
35 maximum point having the largest similarity degree before weighting
26
is extracted (step S205: NO), the CPU 47 adopts the maximum point
extracted in step S204 as the maximum point (target weighted
similarity degree) for calculating the time delay AT of the waveform
data.
5 If a plurality of weighted maximum points having the largest
similarity degree before weighting are extracted (step S205: YES),
the CPU 47 first extracts the maximum point having a frame delay amount
closest to the frame delay amount where the maximum point became
maximum in the previous yarn travelling speed acquiring processing
10 among the extracted weighted maximum points. The CPU 47 adopts the
extracted weighted maximum point as the maximum point (target weighted
similarity degree) for calculating the time delay AT of the waveform
data (step S206). Since the CPU 47 extracts the maximum point for
calculating the delay AT from the plurality of weighted maximum points,
15 the CPU 47 may be referred to be functioning as the travelling
information acquiring section 67 adapted to extract the maximum point
for calculating the delay AT.
Next, the processing carried out after the adoption
determination processing is terminated will be described. After the
20 execution of the adoption determination processing is terminated,
the CPU 47 returns to the flow of FIG. 9 and proceeds to processing
of step S109.
In step S109, the CPU 47 calculates the travelling speed of the
spun yarn 20 based on the frame delay amount of the maximum point
25 adopted in the adoption determination processing of step S108. In
the following description, the frame delay amount of the maximum point
adopted in step S108 is referred to as "currently adopted maximum
value correspondence delay amount". The currently adopted maximum
value correspondence delay amount is assumed to correspond to the
30 time delay AT between the waveform data of the upstream ring buffer
56 and the waveform data of the downstream ring buffer 55. The
currently adopted maximum value correspondence delay amount is the
frame delay amount where the waveform of the upstream ring buffer
56 and the waveform of the downstream ring buffer 55 match when the
35 downstream frame illustrated in FIG. 13 is moved in the past direction
27
from the most recent position on the time axis of the upstream frame.
FIG. 13 illustrates an example in which the waveform data of the
upstream ring buffer 56 and the waveform data of the downstream ring
buffer 55 match when the downstream frame is shifted by eight indices
5 from index [127] to index [119] (i.e., a case in which the frame delay
amount is eight).
Therefore, in the example of FIG. 13, the time duration from
index [127] to index [119] corresponds to the time delay AT of the
waveform data of the upstream ring buffer 56 and the waveform data
10 of the downstream ring buffer 55.
The CPU 4 7 measures the sampling time interval when sampling
the yarn thickness unevenness signal for every sampling interval.
Since the CPU 47 measures the sampling time interval, the CPU 47
functions as the measuring section 73. The CPU 47 integrates each
15 sampling time interval from index [127] to index [119]. In other words,
the CPU 47 integrates the sampling time interval from index [127]
to index [126], the sampling time interval from index [126] to index
[125], ... and the sampling time interval from index [120] to index
[119]. The integration result becomes the time delay AT of the
20 waveforms.
The effects of calculating the time delay AT of the waveforms
by integrating the sampling time interval will be described. For
example, the sampling time interval becomes constant if the travelling
speed of the spun yarn 20 is constant, and hence the time delay AT
25 of the waveforms is obtained by multiplying the sampling time interval
and the frame delay amount (number of indices). If the average
travelling speed of the spun yarn 20 is in acceleration, the sampling
time interval gradually becomes shorter, and hence the time delay
AT of the waveforms becomes longer than a value obtained by multiplying
30 the most recent sampling time interval and the frame delay amount
(number of indices) . If the average travelling speed of the spun yarn
20 is in deceleration, the time delay AT of the waveforms becomes
shorter. By integrating the sampling time intervals to calculate the
time delay AT of the waveforms as in the present embodiment, the delay
35 AT can be accurately calculated even if the spun yarn 20 is in
28
acceleration or in deceleration.
The CPU 47 calculates the travelling speed of the spun yarn 20
by substituting AT obtained as described above to the equation (1)
(step S109). By calculating the travelling speed of the spun yarn
5 20 in accordance with the weighted maximum point, the CPU 47 can
accurately calculate the travelling speed of the spun yarn 20 even
if a plurality of raw maximum points exists . In the present embodiment,
as described above, the adoption determination processing is
performed for preventing a maximum point having low reliability from
10 being adopted. The CPU 47 thus can obtain highly reliable travelling
speed.
The CPU 47 changes the sampling period (cycle) of the second
A/D converter 4 6 in accordance with the yarn travelling speed obtained
as described above. Specifically, the CPU 47 generates the fixed yarn
15 length pulse signal at a frequency proportional to the yarn travelling
speed, and transmits the generated fixed yarn length pulse signal
to the second A/D converter 46. The yarn travelling speed is an
accurate yarn travelling speed obtained in accordance with the
weighted maximum point. Therefore, by sampling the yarn thickness
20 unevenness signals by the second A/D converter 4 6 in accordance with
the fixed yarn length pulse signal based on the yarn travelling speed,
the number of data per unit length of the spun yarn 20 can be accurately
made constant.
The yarn travelling speed obtained as described above is
25 transmitted to the unit control section 50. The unit control section
50 transmits the control signal to the motor control section 54
according to the travelling speed of the spun yarn 20 transmitted
from the clearer 15, and controls the rotation of the winding drum
24. The winder unit 10 thus can perform the winding of the package
30 30 according to the accurate travelling speed of the spun yarn 20.
The unit control section 50 can calculate a total length of the spun
yarn 20 wound into the package 30 by integrating the travelling speed
of the spun yarn 20 by time. Therefore, for example, when the winding
of the spun yarn 20 of a prescribed length is finished, the winder
35 unit 10 can terminate the winding of the spun yarn 20 and have a
29
fully-wound package. Since the winder unit 10 thus can make the length
of the spun yarn 20 to be wound to the respective package 30 to be
uniform, packages 30 wound with spun yarn 20 of uniform length can
be produced.
5 In the above description, although the CPU 47 has been described
to calculate the travelling speed of the spun yarn 20, the yarn
travelling information is not necessarily acquired in the form of
speed. For example, if the spun yarn 20 moved 2 cm in one second and
3 cm in the next second, the CPU 47 merely needs to acquire the
10 information "moved total of 5 cm in two seconds", and may not
necessarily need to acquire information in the form of speed such
as "2.5 cm per second". The information of "length the spun yarn 20
moved per unit time" is also information relating to the travelling
state of the spun yarn 20, and hence is one type of yarn travelling
15 information.
Next, the method for determining the weighting curve will be
described. In the present embodiment, the weighting curve includes
a history weighting curve and a reference speed weighting curve.
First, the history weighting curve will be described. The
20 history weighting curve is set so that the weighting factor becomes
large in proximity to the frame delay amount where the weighted maximum
point became maximum in the previous yarn travelling speed acquiring
processing. In other words, the maximxom point of the history
weighting curve is set in accordance with the previously calculated
25 frame delay amount. The history weighting curve becomes a weighting
curve corresponding to the speed ratio based on the frame delay amount
(prescribed delay amount) of the first yarn thickness unevenness
signal and the second yarn thickness unevenness signal.
The weighting curve corresponding to the speed ratio will be
30 hereinafter described using specific examples. Suppose that the
first A/D converter 45 samples the yarn thickness unevenness signal
every time the spun yarn 20 travels 1 mm. In this case, if the
travelling speed of the spun yarn 20 calculated in accordance with
the rotational information of the winding drum 24 is 1000 mm/sec,
35 the sampling time interval is 1 msec. In this case, if the distance
30
between the first yarn unevenness detecting sensor 43 and the second
yarn unevenness detecting sensor 44 is 10 mm and the actual travelling
speed of the spun yarn 20 is 1000 mm/sec, the value of the frame
delay amount of the maximum point adopted in step S108 of FIG. 9 becomes
5 ten. However, if there is error in the sampling frequency and the
like, and the calculated frame delay amount is nine, the actual
travelling speed of the spun yarn 20 becomes 1000 x (10/9) mm/sec.
Similarly, if the calculated frame delay amount is 11, the actual
travelling speed of the spun yarn 20 is 1000 x (10/11) mm/sec. Thus,
10 the speed ratio differs even if the frame delay amount is one.
Therefore, the weight corresponding to the speed ratio based on the
frame delay amount needs to be determined, instead of determining
the weight based on the frame delay amount.
As illustrated in FIG. 14A, the CPU 47 obtains the speed ratio
15 in accordance with the frame delay amount to be the reference (a
reference delay amount) and the calculated frame delay amount (a
calculated delay amount) . In FIG. 14A, the numerator represents the
reference delay amount, and the denominator represents the calculated
delay amount. The reference delay amount is the frame delay amount
20 corresponding to the sampling speed, and is ten herein. This speed
ratio is obtained by having the reference delay amount of ten as the
numerator and changing the calculated delay amount of the denominator
one by one. As illustrated in FIG. 14B, the CPU 47 then obtains a
first history correction speed ratio that takes into consideration
25 the frame delay amount previously calculated with respect to the
obtained speed ratio. For example, if the previously calculated
frame delay amount is 13, the first history correction speed ratio
is calculated using the speed ratio (10/13) in a frame A illustrated
in FIG. 14A. Specifically, when obtaining the first history
30 correction speed ratio in a frame C illustrated in FIG. 14B, the CPU
47 calculates the first history correction speed ratio (13/11) in
the frame C by dividing the speed ratio (10/11) of a frame B by the
speed ratio (10/13) of the frame A. Similarly, the first history
correction speed ratio in a frame D can be obtained by dividing the
35 speed ratio of the frame A by the speed ratio of the frame A.
31
As illustrated in FIG. 14C, the CPU 47 then obtains a second
history correction speed ratio. The second history correction speed
ratio is obtained, such that the first history correction speed ratio
(the speed ratio 13/13 of the frame D) corresponding to the previously
5 calculated frame delay amount takes a largest value, by correcting
the first history correction speed ratio exceeding such value.
Specifically, the second history correction speed ratio is calculated
by interchanging the denominator and the numerator of each value of
"13/7", "13/8", ..., "13/12" in which the first history correction
10 speed ratio exceeds 13/13. The first history correction speed ratio
that does not exceed 13/13 is used as it is as the second history
correction speed ratio.
The CPU 47 obtains the history weighting curve using the second
history correction speed ratio. Specifically, the history weighting
15 curve is defined with the following equation (4) designating the
relationship of a history weighting factor Wp that forms the history
weighting curve and a second history correction speed ratio N.
Wp = exp(-(1-N) VW) ••• (4)
W is a constant that can be set by the user.
20 FIG. 14D illustrates an example of the history weighting curve
calculated using the equation (4). The history weighting curve
having the previous frame delay amount of 13 as maximum is illustrated.
The above-described example illustrates an example of
calculating the history weighting curve when the previously
25 calculated delay amount is 13, but the history weighting curve when
the previously calculated delay amount takes other values is also
calculated in advance. When weighting the maximum point using the
history weighting curve, the CPU 47 can select and use the history
weighting curve corresponding to the previous calculated delay
30 amount.
Next, the sampling speed weighting curve will be described. The
sampling speed weighting curve is set such that the weighting factor
becomes large in proximity to the frame delay amount corresponding
to the sampling speed calculated in accordance with the rotational
35 information of the winding drum 24, as described above. The maximum
32
point of the sampling speed weighting curve is set in accordance with
the frame delay amount when the spun yarn 20 travelled under the
sampling speed. Similar to the history weighting curve, regarding
the sampling speed weighting curve, the weight is not determined based
5 on the frame delay amount but rather the weight corresponding to the
speed ratio based on the frame delay amount needs to be determined.
Therefore, the sampling speed weighting curve becomes a weighting
curve corresponding to the speed ratio based on the frame delay amount
(prescribed delay amount) of the first yarn thickness unevenness
10 signal and the second yarn thickness unevenness signal.
As illustrated in FIG. 15A, the CPU 47 obtains the speed ratio
in accordance with the frame delay amount to become the reference
(the reference delay amount) and the calculated frame delay amount
(the calculated delay amount) . In FIG. 15A, the numerator represents
15 the reference delay amount, and the denominator represents the
calculated delay amount. The reference delay amount is the frame
delay amount corresponding to the sampling speed, and is ten herein.
This speed ratio is obtained by having the reference delay amount
of ten as the numerator and changing the calculated delay amount of
20 the denominator one by one.
The CPU 47 then obtains the sampling speed correction speed
ratio, as illustrated in FIG. 15B. The sampling speed correction
speed ratio is obtained, such that the speed ratio (a speed ratio
10/10 of a frame E) corresponding to the frame delay amount when the
25 spun yarn 20 travelled under the sampling speed takes a largest value,
by correcting the speed ratio exceeding such value. Specifically,
the sampling speed correction speed ratio is calculated by
interchanging the denominator and the numerator of each value of
"10/7", "10/8", "10/9" in which the speed ratio exceeds 10/10. The
30 correction speed ratio not exceeding 10/10 is used as it is as the
sampling speed correction speed ratio.
The CPU 47 uses the sampling speed correction speed ratio to
obtain the sampling speed weighting curve. Specifically, the
sampling speed weighting curve is defined with the following.equation
35 (5) designating a relationship of a sampling speed weighting factor
33
WA that forms the sampling speed weighting curve and a sampling speed
correction speed ratio N.
Wa = exp(-(l-N) VW) ••• (5)
W is a constant that can be set by the user.
5 FIG. 15C illustrates an example of a sampling speed weighting
curve calculated using the equation (5). An example illustrated in
FIG. 15C is a sampling speed weighting curve in which the frame delay
amount (10) when the spun yarn 20 travelled under the sampling speed
is maximum.
10 The above-described example illustrates an example of
calculating the sampling speed weighting curve when the spun yarn
20 travelled under the sampling speed and the frame delay amount is
ten. However, since the sampling speed varies, the sampling speed
weighting curve is calculated in advance for every sampling speed.
15 When weighting the maximum point using the sampling speed weighting
curve, the CPU 47 can select and use the sampling weighting curve
of the frame delay amount corresponding to the sampling speed.
The CPU 47 can correct the history weighting factor Wp of the
history weighting curve and the sampling speed weighting factor WA
20 of the sampling speed weighting curve in accordance with the weighted
similarity degree of when the weighted maximum point becomes maximum
in the previous yarn travelling speed acquiring processing.
Specifically, provided that the weighted similarity degree is Sp when
the weighted maximum point becomes maximum in the previous yarn
25 travelling speed acquiring processing, the corrected history
weighting factor Wp' and the corrected sampling speed weighting factor
WA' are defined with the following equations (6) and (7).
Wp' = Sp X Wp ••• (6)
WA' = (2S' - SP)WA— (7)
30 S' is a stable similarity factor, and a minimum value of the similarity
degree, and the like, for example, can be set.
The correction of the history weighting factor and the sampling
speed weighting factor emphasizes the influence of the history
weighting factor Wp when the similarity degree Sp is large and
35 emphasizes the influence of the sampling speed weighting factor when
34
the similarity degree Sp is small. This is based on the fact that
if the previous similarity degree Sp is large, the history weighting
curve is assumed to have higher reliability than the sampling speed
weighting curve. Therefore, the maximum point having a high
5 reliability can be selected by performing the correction. Since the
CPU 47 corrects the history weighting factor and the sampling speed
weighting factor in accordance with the previous similarity degree,
the CPU 47 thus also functions as the travelling information acquiring
section 67 for correcting the history weighting factor and the
10 sampling speed weighting factor.
The present embodiment is configured as described above, and
the similarity degree evaluating section 65 creates the first
imaginary frame in accordance with the first yarn thickness unevenness
signal (first extracted signal) extracted with the first extracting
15 condition from the first yarn thickness unevenness signal (first
waveform data series), and creates the second imaginary frame in
accordance with the second yarn thickness unevenness signal (second
extracted signal) extracted with the second extracting condition from
the second yarn thickness unevenness signal (second waveform data
20 series). The similarity degree evaluating section 65 obtains a
plurality of similarity degrees of the first yarn thickness unevenness
signal (first extracted signal) and the second yarn thickness
unevenness signal (second extracted signal) based on the first
imaginary frame and the second imaginary frame. Therefore, by
25 extracting the signals from the first yarn thickness unevenness
signals (first waveform data series) and the second yarn thickness
unevenness signals (second waveform data series), the number of
signals of the first yarn thickness unevenness signals (first waveform
data series) and the second yarn thickness unevenness signals (second
30 waveform data series) within the first imaginary frame and the second
imaginary fame can be reduced. The processing load in calculating
the similarity degree can be reduced by obtaining the similarity
degree using the first imaginary frame and the second imaginary frame
in which the number of signals is reduced. Therefore, even if the
35 acquiring number of the first yarn thickness unevenness signals (first
35
waveform data series) and the second yarn thickness unevenness signals
(second waveform data series) is increased to increase the resolution,
the load in calculating the similarity degree can be reduced.
The similarity degree evaluating section 65 changes the second
5 extracting condition for extracting the waveform data that configures
the upstream frame and the first extracting condition for extracting
the waveform data that configures the downstream frame in accordance
with the sampling speed. The more appropriate upstream frame and the
downstream frame corresponding to the sampling speed thus can be
10 obtained. The similarity degree thus can be more appropriately
calculated.
The weighting processing section 66 performs the weighting
processing on each of a plurality of maximum points using the weighting
curve (the history weighting curve and the sampling speed weighting
15 curve) corresponding to the speed ratio in accordance with the frame
delay amount of the first yarn thickness unevenness signal (first
extracted signal) forming the first imaginary frame and the second
yarn thickness unevenness signal (second extracted signal) forming
the second imaginary frame. The weight can be appropriately set for
20 every frame delay amount by using the weighting curve corresponding
to the speed ratio based on the frame delay amount. Therefore, by
using the weighting curve corresponding to the speed ratio based on
the delay amount, the weighting processing can be more accurately
performed and the travelling state of the yarn can be more accurately
25 acquired.
The winder unit 10 includes the clearer 15 capable of
calculating the travelling speed of the spun yarn 10, as described
above, and thus performs a control of each section using the accurate
travelling information of the spun yarn 10 acquired by the clearer
30 15.
One embodiment of the present invention has been described above,
but the present invention is not limited to the above embodiment.
For example, the spun yarn 20 is traversed on the surface of the package
30 while rotating the package 30 with the rotating winding drum 24
35 in the embodiment described above, but this is not the sole case,
36
and the configuration of the present invention can be applied even
with the yarn winding machine having a configuration in which the
drive of the package and the traverse operation are independent. Such
a yarn winding machine includes an automatic winder provided with
5 an arm-type traverse device that traverses the spun yarn 20 with a
swinging arm, and a belt-type traverse device that traverses the spun
yarn 20 with a yarn hooking member that reciprocates to the left and
the right by a belt.
The yarn travelling information acquiring device according to
10 an embodiment of the present invention is not limited to being provided
in the automatic winder, and may be provided in other yarn processing
devices such as a fine spinning machine, and the like, for example.
In the above embodiment, the change in the light receiving
amount is monitored by the yarn unevenness detecting sensors 43 and
15 44, but for example, a yarn unevenness detecting sensor of a type
that detects the change in electrostatic capacitance of the travelling
spun yarn 20 may be adopted. With such a configuration, the change
in mass per unit length of the spun yarn 20 can be detected. In other
words, the yarn unevenness detecting sensor is sufficient to be
20 configured to detect the thickness unevenness of the spun yarn 20
by some kind of method.
The weighting curve does not need to be in a form of exponential
function as described in the embodiment, and merely needs to be able
to perform weighting in some form.
25 In the embodiment described above, the weighting curve (the
history weighting curve and the sampling speed weighting curve) is
defined as the exponential function (exp function), but the present
invention is not limited thereto, and any function can be used for
the definition of the weighting curve as long as the function has
30 one maximum value.
Description has been made that the clearer 15 acquires the yarn
travelling information such as the fixed yarn length pulse signal
and the yarn travelling speed, but instead, the clearer 15 may be
configured to acquire other yarn travelling information. For example,
35 the clearer 15 may obtain the total length of the traveled spun yarn
37
20 by integrating the obtained yarn travelling speed by time.
In the clearer 15, for example, only the time delay AT of the
waveforms may be obtained without calculating the yarn travelling
speed. Since the time delay AT of the waveform generates when the
5 spun yarn 20 travels, the time delay AT can be referred to as the
travelling information of the spun yarn 20. In this case, the time
delay AT of the waveform obtained by the clearer 15 is output to the
unit control section 50, and calculation of the yarn travelling speed
using AT can be carried out by the unit control section 50.
10 In the above embodiment, the yarn travelling speed calculated
by the clearer 15 is output to the unit control section 50, but the
yarn travelling speed may be output as numerical data or may be output
in other forms. For example, the fixed yarn length pulse signal
described above may be output to the unit control section 50.
15 Since the second A/D converter 46 arranged in the clearer 15
is provided for performing the FFT calculation, the second A/D
converter 46 may be omitted if the FFT calculation is not carried
out.
In the above embodiment, the sampling speed is acquired in
20 accordance with the rotation pulse signal from the rotation sensor
42 configured as a rotary encoder, but the present invention is not
limited thereto.
The above embodiment describes an example in which the history
weighting curve and the sampling speed weighting curve are used as
25 the weighting curve, but only either one of the weighting curves may
be used.
In the above embodiment, the head position of the downstream
frame is fixed, and the similarity degree is obtained while shifting
the head position of the upstream frame. In place of this
30 configuration, the head position of the upstream frame may be fixed,
and the similarity degree may be obtained while shifting the head
position of the downstream frame. The similarity degree may be
obtained while shifting both the downstream frame and the upstream
frame. However, if the downstream frame is fixed at a position
35 including the most recent data of the waveform data included in the
38
downstream ring buffer, the similarity degree can be calculated in
real time while always using the most recent data.
In the above embodiment, the functions of the similarity degree
evaluating section 65, the weighting processing section 66, the
5 travelling information acquiring section 67, the yarn quality
measuring section 68, the sampling speed acquiring section 72, the
measuring section 73, and the like are realized by hardware and
software, but some or all of these functions may be realized with
a dedicated hardware.
10 A yarn travelling information acquiring device according to an
embodiment of the present invention includes a first detecting section,
a second detecting section, a similarity degree evaluating section,
and a travelling information acquiring section. The first detecting
section is adapted to detect a thickness unevenness of a travelling
15 yarn and to output first yarn thickness unevenness signals. The
second detecting section is arranged at a predetermined distance from
the first detecting section in a yarn travelling direction, and is
adapted to detect the thickness unevenness of the yarn and to output
second yarn thickness unevenness signals. The similarity degree
20 evaluating section is adapted to use a first imaginary frame and a
second imaginary frame to select a plurality of positions of the first
imaginary frame on a time axis of the second imaginary frame within
a second time range, and to evaluate a plurality of similarity degrees
of the first yarn thickness unevenness signals and the second yarn
25 thickness unevenness signals. The first imaginary frame is formed
in accordance with the first yarn thickness unevenness signals
extracted from the first yarn thickness unevenness signals acquired
within a predetermined first time range. The second imaginary frame
is formed in accordance with the second yarn thickness unevenness
30 signals extracted from the second yarn thickness unevenness signals
acquired within the second time range that is longer than the first
time range. The travelling information acquiring section is adapted
to calculate a time delay between the first yarn thickness unevenness
signals and the second yarn thickness unevenness signals in accordance
35 with the similarity degrees, and to acquire travelling information
39
of the yarn in accordance with the predetermined distance and the
time delay.
Therefore, since the signals are extracted from the first yarn
thickness unevenness signals and the second yarn thickness unevenness
5 signals to create the first imaginary frame and the second imaginary
frame, the number of signals of the first yarn thickness unevenness
signals and the second yarn thickness unevenness signals within the
first imaginary frame and the second imaginary fame can be reduced.
The processing load in calculating the similarity degree can be
10 reduced by obtaining the similarity degree using the first imaginary
frame and the second imaginary frame in which the number of signals
is reduced. Therefore, even if the acquiring number of the first yarn
thickness unevenness signals and the second yarn thickness unevenness
signals is increased to increase the resolution, the load in
15 calculating the similarity degree can be reduced. Accordingly, the
travelling state of the yarn can be accurately acquired while reducing
the processing load in the yarn travelling information acquiring
device according to an embodiment of the present invention.
The yarn travelling information acquiring device further
20 includes a sampling speed acquiring section adapted to acquire a
sampling speed of the yarn. The similarity degree evaluating section
is adapted to change according to the sampling speed at least one
of a first extracting condition for extracting the first yarn
thickness unevenness signals that form the first imaginary frame and
25 a second extracting condition for extracting the second yarn thickness
unevenness signals that form the second imaginary frame. In this case,
a more appropriate first imaginary frame and/or second imaginary frame
according to the sampling speed can be obtained. The similarity
degree thus can be more appropriately calculated.
30 The yarn travelling information acquiring device further
includes a weighting processing section. The weighting processing
section is adapted to carry out a weighting processing using a
weighting factor designated by a weighting curve according to a
reference delay amount of the.first yarn thickness unevenness signal
35 and the second yarn thickness unevenness signal, and a speed ratio,
40
which is a ratio of the calculated delay amount with respect to the
reference delay amount for a plurality of similarity degrees obtained
by the similarity degree evaluating section, and to calculate a
plurality of weighted similarity degrees. The travelling
5 information acquiring section is adapted to acquire the travelling
information in accordance with the weighted similarity degrees on
which the weighting processing has been carried out by the weighting
processing section. Through the use of the weighting curve
corresponding to the speed ratio, the weighting processing can be
10 more accurately performed and the travelling state of the yarn can
be more accurately acquired.
The yarn processing device preferably includes the yarn
travelling information acquiring device described above, a yarn
processing section adapted to perform processing on the yarn, and
15 a control section adapted to control the processing performed by the
yarn processing section based on the travelling information of the
yarn acquired by the yarn travelling information acquiring device.
In this case, the yarn processing section can be controlled using
accurate travelling information of the yarn acquired by the yarn
20 travelling information acquiring device.
According to the present invention, the travelling state of the
yarn can be accurately acquired while reducing the processing load.

WE CLAIM:
1. A yarn travelling information acquiring device comprising:
a first detecting section adapted to detect a thickness
unevenness of a travelling yarn and to output first yarn thickness
5 unevenness signals;
a second detecting section arranged upstream in a yarn
travelling direction at a distance from the first detecting section,
and adapted to detect the thickness unevenness of the yarn and to
output second yarn thickness unevenness signals;
10 a similarity degree evaluating section adapted to use a first
imaginary frame and a second imaginary frame to select a plurality
of reference positions of the first imaginary frame on a time axis
of the second imaginary frame, and to evaluate a plurality of
similarity degrees of the first yarn thickness unevenness signals
15 and the second yarn thickness unevenness signals, the first imaginary
frame being formed in accordance with the first yarn thickness
unevenness signals extracted from the first yarn thickness unevenness
signals acquired within a first time range, and the second imaginary
frame being formed in accordance with the second yarn thickness
20 unevenness signals extracted from the second yarn thickness
unevenness signals acquired within a second time range that is longer
than the first time range; and
a travelling information acquiring section adapted to calculate
a time delay between the first yarn thickness unevenness signals
25 forming the first imaginary frame and the second yarn thickness
unevenness signals forming the second imaginary frame in accordance
with the similarity degrees, and to acquire travelling information
of the yarn in accordance with the distance and the time delay.
2. The yarn travelling information acquiring device according
to claim 1, further comprising a sampling speed acquiring section
adapted to acquire a sampling speed of the yarn,
wherein the similarity degree evaluating section is adapted to
change according to the sampling speed at least one of a first
35 extracting condition for extracting the first yarn thickness
42
unevenness signals that form the first imaginary frame and a second
extracting condition for extracting the second yarn thickness
unevenness signals that form the second imaginary frame.
5 3. The yarn travelling information acquiring device according
to claim 2, wherein
when the sampling speed acquired by the sampling speed acquiring
section is less than a threshold, the similarity degree evaluating
section is adapted to change the first extracting condition and the
10 second extracting condition so as to not extract the first yarn
thickness unevenness signals output by the first detecting section
and the second yarn thickness unevenness signals output by the second
detecting section, and
when the sampling speed acquired by the sampling speed acquiring
15 section is equal to or greater than the threshold, the similarity
degree evaluating section is adapted to change the first extracting
condition and the second extracting condition so as to extract the
first yarn thickness unevenness signals output by the first detecting
section and the second yarn thickness unevenness signals output by
20 the second detecting section.
4. The yarn travelling information acquiring device according
to any one of claim 1 through claim 3, further comprising a weighting
processing section adapted to carry out a weighting processing on
25 the plurality of similarity degrees obtained by the similarity degree
evaluating section using a weighting factor designated by a weighting
curve according to a speed ratio based on a delay amount of the first
yarn thickness unevenness signals forming the first imaginary frame
and the second yarn thickness unevenness signals forming the second
30 imaginary frame, and to calculate a plurality of weighted similarity
degrees,
wherein the travelling information acquiring section is adapted
to acquire the travelling information in accordance with the weighted
similarity degrees on which the weighting processing has been carried
35 out by the weighting processing section.
43
5. The yarn travelling information acquiring device according
to any one of claim 1 through claim 4, wherein the travelling
information acquiring section is adapted to acquire as the travelling
5 information of the yarn, at least one of travelling speed of the yarn,
a fixed-length pulse of the yarn, a moved length of the yarn per unit
time, and a time delay of a waveform of the first yarn thickness
unevenness signals forming the first imaginary frame and a waveform
of the second yarn thickness unevenness signals forming the second
10 imaginary frame.
6. A yarn processing device comprising:
the yarn travelling information acquiring device according to
any one of claim 1 through claim 5;
15 a yarn processing section adapted to perform a processing on
the yarn; and
a control section adapted to control the processing of the yarn
processing section in accordance with the travelling information of
the yarn acquired by the yarn travelling information acquiring device .
20
7. A yarn travelling information acquiring method comprising
the following steps:
detecting a thickness unevenness of a travelling yarn by a first
detecting section and generating first yarn thickness unevenness
25 signals;
detecting the thickness unevenness of the yarn by a second
detecting section located upstream in a yarn travelling direction
at a distance from the first detecting section, and generating second
yarn thickness unevenness signals;
30 using a first imaginary frame and a second imaginary frame,
selecting a plurality of reference positions of the first imaginary
frame on a time axis of the second imaginary frame, and evaluating
a plurality of similarity degrees of the first yarn thickness
unevenness signals and the second yarn thickness unevenness signals,
35 forming the first imaginary frame in accordance with the first yarn
44
thickness unevenness signals extracted from the first yarn thickness
unevenness signals acquired within a first time range, and forming
the second imaginary frame in accordance with the second yarn
thickness unevenness signals extracted from the second yarn thickness
5 unevenness signals acquired within a second time range that is longer
than the first time range; and
calculating a time delay between the first yarn thickness
unevenness signals forming the first imaginary frame and the second
yarn thickness unevenness signals forming the second imaginary frame
10 in accordance with the similarity degrees, and generating travelling
information of the yarn in accordance with the distance and the time
delay.
8. The yarn travelling information acquiring method according
15 to claim 1, further comprising the steps of:
acquiring a sampling speed of the yarn,
changing according to the sampling speed at least one of a first
extracting condition for extracting the first yarn thickness
unevenness signals that form the first imaginary frame and a second
20 extracting condition for extracting the second yarn thickness
unevenness signals that form the second imaginary frame.
9. The yarn travelling information acquiring method according
to claim 8, wherein
25 when the sampling speed acquired is less than a threshold,
changing the first extracting condition and the second extracting
condition so as to not extract the first yarn thickness unevenness
signals and the second yarn thickness unevenness signals, and
when the sampling speed acquired is equal to or greater than
30 the threshold, changing the first extracting condition and the second
extracting condition so as to extract the first yarn thickness
unevenness signals and the second yarn thickness unevenness signals.
10. The yarn travelling information acquiring method
35 according to any one of claim 7 through claim 9, further comprising
45
the step of:
weighting the plurality of similarity degrees obtained using
a weighting factor designated by a weighting curve according to a
speed ratio based on a delay amount of the first yarn thickness
5 unevenness signals forming the first imaginary frame and the second
yarn thickness unevenness signals forming the second imaginary frame,
and calculating a plurality of weighted similarity degrees, and
acquiring the travelling information in accordance with the
weighted similarity degrees on which the weighting processing has
10 been carried out.
11. The yarn travelling information acquiring method
according to any one of claim 7 through claim 10, further comprising
the step of acquiring as the travelling information of the yarn, at
15 least one of travelling speed of the yarn , a fixed-length pulse of
the yarn, a moved length of the yarn per unit time, and a time delay
of a waveform of the first yarn thickness unevenness signals forming
the first imaginary frame and a waveform of the second yarn thickness
unevenness signals forming the second imaginary frame.