[Technical Field]

The present invention relates to a plant state
monitoring system and a plant state monitoring method for
monitoring plant states of water and sewage plants,
chemical plants, and power plants.
[Background Art]

Since 2011, revenues of water rates have declined due
to depopulation and water saving, and a financial situation
is severe. The number of employees involved in the water
and sewages has been decreased. Particularly, in recent
years, skilled employees who have plentifully acquired
know-hows and skills reach their mandatory retirement age,
and thus, there is a shortage of employee. As a result,
the know-hows and skills have not been transferred. This
trend is not limited to a water supply business, and a
chemical industry and an electric power industry have
similar problems.
[0003]
The skilled employee can detect a state change of the
plant (facility) by listening to sound generated by a pump
or a motor, smelling slight chemical odors and ozone odors,
and even looking at the facility. The detection of the
3
state change based on such environment data greatly relies
on senses, and thus, it is hard to leave the detection of
the state change in a manual.
[0004]
However, as the number of skilled employees is
decreased, such know-hows and skills may be lost. Thus,
systems and methods for replacing these know-hows and
skills are required.
[0005]
JP 2002-041143 A (Patent Document 1) is a background
art of the present technical field. Patent Document 1
describes an abnormality diagnosis method of an operation
unit of recording an acoustic signal acquired by a
microphone in a data recorder, acquiring one or more
filter-processed signals in which one or more types of
acoustic waveforms corresponding to a frequency component
of sound output from at least one operation unit of a
plurality of operation units clearly appeared by performing
filter processing on the acoustic signal in a filter
device, and determining whether or not abnormality occurs
in each operation unit based on the acoustic waveform in a
data analyzer (see Abstract).
[Prior Art Document]
[Patent Document]
[0006]
[Patent Document 1] JP 2002-041143 A
[Summary of the Invention]
4
[Problem to be Solved by the Invention]
[0007]
Patent Document 1 describes the abnormality diagnosis
method of an operation unit capable of individually
diagnosing the abnormalities of the plurality of operation
units based on the acoustic signals. Patent Document 1
describes that a “timing on a time axis of sound generated
from each of the plurality of operation units is specified
in advance” (see 0010).
[0008]
That is, the abnormality diagnosis method of an
operation unit described in Patent Document 1 specifies the
operation unit from which the sound is generated by
specifying the timing of the sound generated from each of
the plurality of operation units in advance, and diagnoses
the abnormality of the operation unit from which the sound
is generated.
[0009]
However, since the abnormality diagnosis method of an
operation unit described in Patent Document 1 is necessary
to specify the timing of the sound generated from each of
the plurality of operation units in advance, that is, is
necessary to prior scheduling of the operation unit, there
is a possibility that it is difficult to acquire the sound
in a non-steady state.
[0010]
In Patent Document 1, when the operation unit is
5
different, since it is necessary to acquire the timing of
the sound and the sound in a steady state again, a case
where the operation unit is different is not described.
[0011]
Therefore, the present invention provides a plant
state monitoring system and a plant state monitoring method
capable of efficiently saving environment data for
detecting a state change of a monitored facility by using
operation data without distinguishing between a steady
state or a non-steady state with no prior scheduling.
[Means for solving the Problems]
[0012]
In order to solve the aforementioned problems, a
plant state monitoring system according to the present
invention includes an environment data measurement device
that acquires environment data of a monitored facility, a
monitoring control device that controls the monitored
facility, acquires operation data of the monitored
facility, acquires the environment data acquired by the
environment data measurement device, and executes
preprocessing of the environment data, and a data storage
unit that stores the environment data acquired by the
environment data measurement device and the operation data
acquired by the monitoring control device in association
with each other.
[0013]
A plant state monitoring method according to the
6
present invention includes acquiring environment data and
operation data of a monitored facility, and executing
labeling processing for giving a label to the environment
data by using the operation data, and storing the
environment data and the operation data in association with
each other.
[Effect of the Invention]
[0014]
According to the present invention, it is possible to
provide the plant state monitoring system and the plant
state monitoring method capable of efficiently saving the
environment data for detecting the state change of the
monitored facility by using the operation data without
distinguishing between a steady state or a non-steady state
with no prior scheduling.
[0015]
Other objects, configurations, and effects will be
made apparent in descriptions of embodiments to be
described below.
[Brief Description of the Drawings]
[0016]
FIG. 1 is an explanatory diagram for describing a
plant state monitoring system 10 according to a first
embodiment.
FIG. 2 is an explanatory diagram for describing a
monitoring control device 300 according to the first
embodiment.
7
FIG. 3 is an explanatory diagram for describing a
workflow of an environment data preprocessing function 330
according to the first embodiment.
FIG. 4 is an explanatory diagram for describing a
case where a label is given to environment data according
to the first embodiment.
FIG. 5 is an explanatory diagram for describing a
plant state monitoring system 10 according to a second
embodiment.
FIG. 6 is an explanatory diagram for describing a
case where a label is given to environment data according
to the second embodiment.
FIG. 7 is an explanatory diagram for describing
clustering (classification) of the environment data
according to the second embodiment.
FIG. 8 is an explanatory diagram for describing a
monitoring control device 300 according to a third
embodiment.
FIG. 9 is an explanatory diagram for describing a
relationship between a failure occurring around a pump and
operation data or environment data according to the third
embodiment.
FIG. 10 is an explanatory diagram for describing a
case where a label is given to environment data according
to a fourth embodiment.
[Best mode for carrying out the Invention]
[0017]
8
Hereinafter, embodiments will be described with
reference to the drawings. Substantially identical or
similar components are denoted by identical reference
numerals, and the description may be omitted when the
description is redundant.
[First Embodiment]
[0018]
First, a plant state monitoring system 10 according
to a first embodiment will be described.
[0019]
FIG. 1 is an explanatory diagram for describing the
plant state monitoring system 10 according to the first
embodiment.
[0020]
The plant state monitoring system 10 described in the
present embodiment includes an environment data measurement
device 200 that acquires environment data of the monitored
facility 100, a monitoring control device 300 that controls
the monitored facility 100, acquires operation data of the
monitored facility 100, acquires the environment data
acquired by the environment data measurement device 200,
and executes preprocessing of the environment data, a data
storage unit 400 that stores the environment data acquired
by the environment data measurement device 200 and the
operation data acquired by the monitoring control device
300 in association with each other, and can delete the
environment data and the operation data if necessary, and
9
state change detection means 500 for detecting a state
change of the monitored facility 100.
[0021]
The environment data measurement device 200 and the
monitoring control device 300 are connected to the
monitored facility 100, and the data storage unit 400 and
the state change detection means 500 are connected to the
monitoring control device 300. The environment data
measurement device 200 and the monitoring control device
300 are connected, and the data storage unit 400 and the
state change detection means 500 are connected.
[0022]
A network is used for these connections. The network
is the Internet, a wide area network (WAN), or a local area
network (LAN).
[0023]
The monitored facility 100 is, for example, a machine
such as a pump or a motor. The state of the monitored
facility 100 is changed during an operation and/or a
failure. This state change is a change of an operation
state of the monitored facility 100, and is a change
(deviation) from a steady state.
[0024]
The monitored facility 100 is often installed in a
plurality of systems in a plant in many cases, and in such
a case, a plurality of monitored facilities 100 is often
installed in the same room (the same space) in many cases.
10
Thus, even when the monitored facilities 100 such as a
plurality of pumps and motors are simultaneously in
operation, it is necessary to appropriately detect changes
of the states of the monitored facilities 100 such as the
plurality of pumps and motors.
[0025]
In the present embodiment, a description will be made
below in particular of one pump that injects chemicals in a
water supply plant as the monitored facility 100. In the
present embodiment, “sound generated by the pump” is used
as the environment data, and an “On or Off signal of the
pump” is used as the operation data. The environment data
includes electromagnetic waves generated by the pump in
addition to the “sound generated by the pump”, and the
operation data includes a pressure signal of the pump, an
output signal of the pump, a current signal of the pump,
and a flow rate signal of the pump in addition to the “On
or Off signal of the pump”.
[0026]
The environment data measurement device 200 is a
device having a sound collecting function and a sound
recording function for sound (environment data) generated
by the pump (monitored facility 100). For example, the
environment data measurement device is a microphone (sound
collecting device) and a data recorder (sound recording
device) that records the sound acquired by the sound
collecting device. One or two or more environment data
11
measurement devices 200 are installed, and acquire
operation sound and failure sound of the pump as digital
data (environment data).
[0027]
The monitoring control device 300 is a device that
controls an operation of another facility of the water
supply plant including the pump (the monitored facility
100). The monitoring control device is, for example, a
control panel. The monitoring control device 300 acquires
the On or Off signal (operation data) of the pump, and
executes preprocessing of the environment data.
[0028]
The monitoring control device 300 acquires the
operation data of the monitored facility 100, and outputs a
control signal to the monitored facility 100.
[0029]
The data storage unit 400 is a recording medium
mounted on hardware such as a personal computer and a
tablet, and stores the environment data and the operation
data.
[0030]
The state change detection means 500 is a program
recorded on the recording medium, executes data analysis
based on the environment data and/or the operation data,
and detects the state change of the monitored facility 100.
The state change detection means 500 may be installed on
the recording medium of the data storage unit 400, or may
12
be installed on a recording medium different from the
recording medium of the data storage unit 400.
[0031]
Next, the monitoring control device 300 according to
the first embodiment will be described.
[0032]
FIG. 2 is an explanatory diagram for describing the
monitoring control device 300 according to the first
embodiment.
[0033]
The monitoring control device 300 described in the
present embodiment has an operation control function 310, a
measurement data reception function 320, an environment
data preprocessing function 330, a data transmission
function 340, and an alarm output function 350.
[0034]
The operation control function 310 controls the
operation of another facility of the water supply plant
including the pump (monitored facility 100). The operation
control function 310 acquires the operation data of the
monitored facility 100, and outputs a control signal to the
monitored facility 100. For example, the operation control
function 310 acquires the On or Off signal (operation data)
of the pump, and outputs the On or Off signal (control
signal) to the pump.
[0035]
The measurement data reception function 320 acquires
13
the On or Off signal (operation data) of the pump acquired
by the operation control function 310 and the sound
(environment data) generated by the pump acquired by the
environment data measurement device 200 together with a
time. The measurement data reception function 320 may
acquire operation data other than the sound generated by
the pump, for example, operation data (measurement data
which is a result acquired by operating the pump according
to the control signal) such as the pressure signal of the
pump, the current signal of the pump, the flow rate signal
of the pump, and the output signal of the pump.
[0036]
The environment data preprocessing function 330
executes preprocessing of the environment data acquired by
the measurement data reception function 320. The
environment data preprocessing function 330 gives a label
to the environment data by using the On or Off signal
(operation data) of the pump (monitored facility 100) and
the sound (sound information) (environment data) generated
by the pump (monitored facility 100) together with the
time.
[0037]
Here, the preprocessing is to decompose the sound
(environment data) generated by the pump into a frequency,
a sound pressure level, and a frequency component, give a
label to the sound (environment data) generated by the
pump, and remove noise from the sound (environment data)
14
generated by the pump. These processing may be referred to
as “decomposition processing”, “labeling processing”, and
“noise removal processing”, respectively.
[0038]
The data transmission function 340 transmits, to the
data storage unit 400, the environment data and the
operation data acquired by the measurement data reception
function 320 and the environment data after the
preprocessing executed by the environment data
preprocessing function 330. In particular, in the present
embodiment, the preprocessed environment data is
transmitted to the data storage unit 400 by executing the
decomposition processing, executing the labeling processing
by using the On or Off signal (operation data), and
executing the noise removal processing for the sound
(environment data) generated by the pump acquired by the
measurement data reception function 320.
[0039]
The alarm output function 350 receives the state
change of the pump detected by the state change detection
unit 500, and outputs an alarm if necessary.
[0040]
Next, a workflow of the environment data
preprocessing function 330 according to the first
embodiment will be described.
[0041]
FIG. 3 is an explanatory diagram for describing the
15
workflow of the environment data preprocessing function 330
according to the first embodiment.
[0042]
The environment data preprocessing function 330
executes preprocessing in the following steps (S).
[0043]
In S101, a time, operation data of a target pump (On
or Off signal of the pump) and environment data of the
target pump (sound emitted by the pump) are acquired from
the measurement data reception function 320.
[0044]
In S102, acoustic analysis of the environment data
(sound emitted by the pump) is executed. Specifically, the
sound generated by the pump is decomposed into sound
information such as a frequency, a sound pressure level,
and a frequency component by using Fourier transform. That
is, the “decomposition processing” is executed. The sound
information such as the decomposed frequency (for which
acoustic analysis is executed), sound pressure level, and
frequency component is also the environment data.
[0045]
In S103, the label is given to the environment data
(for example, frequency) on which the decomposition
processing is executed by using the On or Off signal of the
target pump. That is, the “labeling processing” is
executed.
[0046]
16
In S104, stationary noise (such as spatial noise) is
removed by using the environment data (for example,
frequency) of the target pump on which the labeling
processing is executed when the target pump is turned off.
That is, the “noise removal processing” is executed.
[0047]
In S105, the preprocessed environment data (for
example, frequency) is transmitted to the data transmission
function 340.
[0048]
Here, although it has been described in the present
embodiment that the environment data (sound generated by
the pump) is decomposed into three types of sound
information of the frequency, the sound pressure level, and
the frequency component, the environment data may be
decomposed into two types of sound information of the
frequency and the sound pressure level, or may be
decomposed into four or more types of sound information.
When it is possible to detect the state change of the
target pump by narrowing down the environment data to sound
information such as a specific (predetermined range)
frequency, sound pressure level, and frequency component
with which the state change of the target pump is easily
detected, the acoustic analysis method is not particularly
limited.
[0049]
In the present embodiment, the preprocessing is
17
executed in the order of the “decomposition processing”,
the “labeling processing”, and the “noise removal
processing”, but the present invention is not necessarily
limited to this order. For example, the decomposition
processing may be executed, the noise removal processing
may be executed, and the labeling processing may be
executed, or the noise removal processing may be executed,
the decomposition processing may be executed, and the
labeling processing may be executed.
[0050]
As described above, a plant state monitoring method
described in the present embodiment acquires the
environment data (sound generated by the pump) and the
operation data (On or Off signal of the pump) of the
monitored facility (pump) 100, and executes the labeling
processing for giving the labels to the environment data by
using the operation data.
[0051]
Next, the labeling of the environment data according
to the first embodiment will be described.
[0052]
FIG. 4 is an explanatory diagram for describing a
case where a label is given to the environment data
according to the first embodiment.
[0053]
The environment data (sound generated by the pump)
acquired by the measurement data reception function 320 is
18
input to the environment data preprocessing function 330
together with the time at regular intervals of about
several minutes. The environment data preprocessing
function 330 continuously acquires the environment data.
[0054]
In the present embodiment, for example, the frequency
is used as the environment data on which the decomposition
processing is executed by the environment data
preprocessing function 330.
[0055]
Meanwhile, the operation data of the target pump (the
On or Off signal of the pump) acquired by the operation
control function 310 is also input to the environment data
preprocessing function 330 together with the time.
[0056]
Since the decomposition processing is executed for
the environment data (for example, frequency), a slight
time delay occurs on the operation data (On or Off signal
of the pump). However, the acquired times of the
environment data and the environment data match each other
based on these times, and the labeling processing is
executed.
[0057]
In the present embodiment, three types of labels are
prepared. A first label is a label when the target pump is
in operation and when the target pump is stopped (no
symbol), a second label is a label (La+ number) when the
19
target pump is in operation in a case where the target pump
is turned on or off, and a third label is a label (Lb +
number) when the target pump is stopped in a case where the
target pump is turned on or off.
[0058]
In the present embodiment, when the target pump is
turned on or off, the label of La or Lb is given to the
environment data for several tens of seconds earlier and
later than the On or Off signal of the target pump. That
is, in the present embodiment, the label (La or Lb) is
given to the environment data earlier and later than the
input of the operation data (the input of the On signal or
the input of the Off signal of the target pump).
[0059]
For example, a label La1 and a label Lb1 are given to
the environment data for several tens of seconds earlier
and later than a signal of Off(1). That is, the label La1
is given to the environment data for several tens of
seconds earlier than the signal of Off(1), and the label
Lb1 is given to the environment data for several tens of
seconds later than the signal of Off(1).
[0060]
A label La2 and a label Lb2 are given to the
environment data for several tens of seconds earlier and
later than a signal of On(1). That is, the label Lb2 is
given to the environment data for several tens of seconds
earlier than the signal of On(1), and the label La2 is
20
given to the environment data for several tens of seconds
later than the signal of On(1).
[0061]
Similarly, a label La3 and a label Lb3, and a label
La4 and a label Lb4 are given to the environment data
earlier and later than a signal of Off(2) and earlier and
later than a signal of On(2).
[0062]
This is because there is a high possibility that the
state of the target pump is changed when the target pump is
turned on or off in this manner.
[0063]
Although it is preferable that such labeling
processing is executed at any time, the labeling processing
may be executed by temporarily storing the environment data
(for example, frequency) and the operation data (On or Off
signal of the pump) in the environment data preprocessing
function 330 and matching the time of the operation data
and the time of the environment data based on these times.
[0064]
In the present embodiment, the environment data (for
example, frequency) and the operation data (On or Off
signal of the pump) of the portion to which the label is
given are stored. That is, the environment data and the
operation data of a portion other than this portion are
deleted if necessary. Accordingly, the environment data
and the operation data can be efficiently saved.
21
[0065]
In particular, in the present embodiment, the noise
removal processing is executed by using the environment
data of the portion of the label when the target pump is
stopped (Off) (for example, frequency), that is, the
environment data of any portion of Lb1, Lb2, Lb3, and Lb4
(for example, frequency).
[0066]
That is, in the present embodiment, the noise removal
processing is executed by removing the environment data to
which the level is given later (meaning of “later” of
earlier and later than the On or Off signal of the target
pump) from the environment data to which the label is given
earlier (meaning of “earlier” of earlier and later than the
On or Off signal of the target pump) or removing the
environment data to which the label is given earlier
(meaning of “earlier” of earlier and later than the On or
Off signal of the target pump) from the environment data to
which the label is given later (meaning of “later” of
earlier and later than the On or Off signal of the target
pump).
[0067]
Accordingly, it is not necessary to newly acquire the
sound of the target pump in order to remove the noise, and
the noise removal processing can be efficiently executed.
[0068]
In the present embodiment, the noise removal
22
processing is executed by using the environment data (for
example, frequency) of the portion of the label when the
target pump is most recently stopped (Off). Accordingly,
it is possible to execute the noise removal processing with
high accuracy.
[0069]
As described above, in this embodiment, it is
possible to efficiently and accurately execute the noise
removal processing by preparing two types of labels (two
types of labels earlier and later than the On or Off signal
of the pump) of the label when the target pump is in
operation or the label when the target pump is stopped for
the On or Off signal of the pump.
[0070]
According to the present embodiment, the environment
data for detecting the state change of the monitored
facility 100 can be efficiently saved by using the
operation data without distinguishing between a steady
state or a non-steady state with no prior scheduling.
[0071]
Although it has been described in the present
embodiment that the On or Off signal of the target pump is
used as the operation data, for example, the output signal
(UP or DOWN signal) of the target pump can be used as the
operation data. Similarly to the On or Off signal of the
target pump, two types of labels are given earlier and
later than the UP or DOWN signal of the target pump.
23
Accordingly, similarly to a case where the On or Off signal
of the target pump is used as the operation data, the
environment data and the operation data can be efficiently
saved, and the noise removal processing can be executed
efficiently and accurately.
[0072]
During a steady operation of the target pump (a state
in which the target pump is steadily in operation other
than during the On or Off or the UP or DOWN), it is
preferable that the environment data is stored for
maintenance by periodically giving the labels.
[0073]
The data storage unit 400 is a database that stores
the environment data (for example, sound generated by the
pump) of the target pump measured by the environment data
measurement device 200, for example, the measurement data
such as the pressure signal of the pump, the current signal
of the pump, the flow rate signal of the pump, and the
output signal of the pump, and the operation data (On or
Off signal of the pump) of the target pump in association
with the time. That is, the data storage unit 400 stores
the environment data and the operation data in association
with each other.
[0074]
The data storage unit 400 secures a recording
capacity of the data storage unit 400 by periodically
deleting these data. In the present embodiment, for
24
example, when one month elapses after the acquisition of
the environment data and the operation data, the recording
capacity of the data storage unit 400 is secured by
deleting the data to which the label (no symbol) is given
during the steady operation or a long term stop of the
target pump.
[0075]
The state change detection means 500 executes the
data analysis on the environment data (for example,
frequency) stored in the data storage unit 400, and detects
the state change of the target pump.
[0076]
As a method of detecting the state change, there are
a physical model, Fault Tree Analysis (FTA), and a
statistical model. In particular, as the statistical
model, there is an adaptive resonance theory (ART) which is
a type of clustering. Through the use of this ART, the
state change of the target pump is detected by extracting
an occurrence pattern of the state change of the target
pump. In the present embodiment, since it is characterized
by efficiently saving the environment data for detecting
the state change of the target pump, the method of
detecting the state change will not be described.
[0077]
The method of detecting the state change is not
limited thereto, and any other detection methods may be
used as long as the state change of the target pump can be
25
detected.
[0078]
When the state change is detected as a result of the
data analysis, the state change detection means 500 outputs
a detection signal of the state change to the alarm output
function 350, and the alarm output function 350 outputs the
alarm. Here, the alarm means that a character is output on
a monitor of a monitoring screen or alarm sound is
produced.
[0079]
Although it has been described in the present
embodiment that the sound generated by the pump is used as
the environment data, the environment data is not limited
to the sound generated by the pump. For example,
electromagnetic waves, light, and odors can be similarly
handled. For example, since the pump generates the
electromagnetic wave, an electromagnetic wave measurement
device that measures the electromagnetic wave can be used
as the environment data measurement device 200, and can be
handled similarly when the pump generates the sound.
[0080]
Although it has been described in the present
embodiment that the pump is used as the monitored facility
100, the monitored facility 100 is not limited to such a
pump. For example, an emergency generator used during a
power failure can be similarly handled. A combustion state
of the emergency generator is measured by using an optical
26
device such as a camera as the environment data measurement
device 200. The environment data preprocessing function
330 decomposes light emitted by the emergency generator
into light information such as luminance and wavelength
(performs the decomposition processing) by using the light
(environment data), and gives the labels by using
appropriate operation data (executes the labeling
processing).
[0081]
Accordingly, similarly to a case where the pump is
used as the monitored facility 100, the environment data
and the operation data can be efficiently saved, and the
noise removal processing can be executed efficiently and
accurately.
[0082]
As described above, the plant state monitoring system
10 described in the present embodiment includes at least
the environment data measurement device 200 that acquires
the environment data of the monitored facility 100, the
monitoring control device 300 that controls the monitored
facility 100, and acquires the operation data of the
monitored facility 100, acquires the environment data
acquired by the environment data measurement device 200,
and executes the preprocessing of the environment data, and
the data storage unit 400 that stores the environment data
acquired by the environment data measurement data 200 and
the operation data acquired by the monitoring control
27
device 300, and is capable of deleting the environment data
and the operation data if necessary.
[0083]
According to the present embodiment, the environment
data for detecting the state change of the monitored
facility 100 can be saved efficiently and selectively by
using the operation data without distinguishing between the
steady state and the non-steady state and with no prior
scheduling.
[0084]
Since the environment data for detecting the state
change of the monitored facility 100 can be selectively
saved, even though the state change is detected by using
environment data having a large intuitive portion, points
are narrowed down, and quality is improved. Accordingly, a
plant state can be monitored.
[0085]
Further, even when the number of skilled employees is
reduced, the system can be unmanned (labor saving) and the
shortage of human resources can be solved by supporting the
monitoring of the plant state as described above.
[Second Embodiment]
[0086]
First, a plant state monitoring system 10 according
to a second embodiment will be described.
[0087]
FIG. 5 is an explanatory diagram for describing the
28
plant state monitoring system 10 according to the second
embodiment.
[0088]
The plant state monitoring system 10 described in the
present embodiment is different from the plant state
monitoring system 10 described in the first embodiment in
that the plant state monitoring system controls a plurality
of monitored facilities (a monitored facility A101, a
monitored facility B102, and monitored facility C103) and
acquires the operation data and the environment data of the
plurality of monitored facilities (the monitored facility
A101, the monitored facility B102, and the monitored
facility C103).
[0089]
That is, the plant state monitoring system 10
described in the present embodiment includes an environment
data measurement device 200 that acquires environment data
of the plurality of monitored facilities (the monitored
facility A101, the monitored facility B102, and the
monitored facility C103), a monitoring control device 300
that controls the plurality of monitored facilities (the
monitored facility A101, the monitored facility B102, and
the monitored facility C103), acquires operation data of
the plurality of monitored facilities (the monitored
facility A101, the monitored facility B102, and the
monitored facility C103), acquires the environment data
acquired by the environment data measurement device 200,
29
and executes preprocessing of the environment data, a data
storage unit 400 that stores the environment data acquired
by the environment data measurement device 200 and the
operation data acquired by the monitoring control device
300, and is capable of deleting the environment data and
the operation data if necessary, and state change detection
means 500 for detecting state changes of the monitored
facilities 100.
[0090]
The monitoring control device 300 described in the
present embodiment has an operation control function 310, a
measurement data reception function 320, an environment
data preprocessing function 330, a data transmission
function 340, and an alarm output function 350.
[0091]
The environment data preprocessing function 330
described in the present embodiment executes the
preprocessing of the environment data acquired by the
measurement data reception function 320 in consideration of
the handling of the plurality of monitored facilities (the
monitored facility A101, the monitored facility B102, and
the monitored facility C103).
[0092]
As in the first embodiment, a case where pumps that
inject chemicals in a water supply plant are used as the
plurality of monitored facilities (the monitored facility
A101, the monitored facility B102, and the monitored
30
facility C103) will be described in the present embodiment.
It is assumed that one pump is installed in each monitored
facility. In the present embodiment, “sound generated by
the pump” is used as the environment data, and an “On or
Off signal of the pump” is used as the operation data.
[0093]
As in the first embodiment, the preprocessing is
“decomposition processing”, “labeling processing”, and
“noise removal processing” in the present embodiment. That
is, the preprocessing is to decompose the sound
(environment data) generated by the pump into a frequency,
a sound pressure level, and a frequency component
(decomposition processing), give a label to the sound
(environment data) generated by the pump (labeling
processing), and remove noise from the sound (environment
data) generated by the pump (noise removal processing).
[0094]
Next, a case where the label is given to the
environment data according to the second embodiment will be
described.
[0095]
FIG. 6 is an explanatory diagram for describing a
case where the label is given to the environment data
according to the second embodiment.
[0096]
As in the first embodiment, the environment data
(sound generated by the pump) acquired by the measurement
31
data reception function 320 is input to the environment
data preprocessing function 330 together with the time at
regular intervals of about several minutes in the present
embodiment. The environment data preprocessing function
330 continuously acquires the environment data.
[0097]
As in the first embodiment, for example, the
frequency is used as the environment data on which the
decomposition processing is executed by the environment
data preprocessing function 330 in the present embodiment.
[0098]
Meanwhile, in the present embodiment, the plurality
of monitored facilities (pumps) is assumed that the
monitored facility B102 is turned on or off while the
monitored facility A101 is in operation and the monitored
facility C103 is turned on while the monitored facility
B102 is in operation.
[0099]
As in the first embodiment, three types of labels are
prepared. A first label is a label when the target pump is
in operation and when the target pump is stopped (no
symbol), a second label is a label (La+ number) when the
target pump is in operation in a case where the target pump
is turned on or off, and a third label is a label (Lb +
number) when the target pump is stopped in a case where the
target pump is turned on or off.
[0100]
32
In the present embodiment, for example, a label La5
and a label Lb5 are given to the environment data for
several tens of seconds earlier and later than a signal of
On(3). That is, the label Lb5 is given to the environment
data for several tens of seconds earlier than the signal of
On(3), and the label La5 is given to the environment data
for several tens of seconds later than the signal of On(1).
[0101]
A label La6 and a label Lb6 are given to the
environment data for several tens of seconds earlier and
later than a signal of On(4). That is, the label Lb6 is
given to the environment data for several tens of seconds
earlier than the signal of On(4), and the label La6 is
given to the environment data for several tens of seconds
earlier than the signal of On(4).
[0102]
A label La7 and a label Lb7 are given to the
environment data for several tens of seconds earlier and
later than a signal of Off(3). That is, the label La7 is
given to the environment data for several tens of seconds
earlier than the signal Off(3), and the label Lb7 is given
to the environment data for several tens of seconds later
than the signal of Off(3).
[0103]
In the present embodiment, the monitored facility
B102 is operated (On) while the monitored facility A101 is
in operation.
33
[0104]
At this time, since the environment data (sound
information (in the present embodiment, for example,
frequency)) of the monitored facility A101 and the
environment data of the monitored facility B102 are
overlapped as the acquired environment data, it is
difficult to separately acquire the environment data of the
monitored facility A101 and the environment data of the
monitored facility B102, and it is difficult to detect the
monitored facility in which the state change occurs.
[0105]
Therefore, the environment data of the monitored
facility B102 can be acquired by using the environment data
of the portion of the label (Lb5) given when the monitored
facility B102 is operated (On) and removing the environment
data of the portion of the label (Lb5) from the environment
data of the portion of the label (Lb5).
[0106]
Similarly, the environment data of the monitored
facility C103 can be acquired by using the environment data
of the portion of the label (Lb6) given when the monitored
facility C103 is operated (On) and removing the environment
data of the portion of the label (Lb6) from the environment
data of the label (La6).
[0107]
Similarly, the environment data of the monitored
facility B102 can be acquired by using the environment data
34
of the portion of the label (Lb7) given when the monitored
facility B102 is stopped (Off) and removing the environment
data of the portion of the label (Lb7) from the environment
data of the portion of the label (La7).
[0108]
That is, in the present embodiment, the noise removal
processing is executed by removing the environment data to
which the level is given “later” from the environment data
to which the label is given “earlier” or removing the
environment data to which the label is given “earlier” from
the environment data to which the label is given “later”.
[0109]
This is because when the monitored facility is turned
on or off, there is a high possibility that the state of
the monitored facility is changed, and the state change of
each of the plurality of monitored facilities in such a
case can be consequently detected.
[0110]
As in the first embodiment, the environment data and
the operation data (On or Off signal of the pump) of the
portion to which the label is given are stored in
association with each other in the data storage unit 400 in
the present embodiment.
[0111]
Next, clustering (classification) of the environment
data according to the second embodiment will be described.
[0112]
35
FIG. 7 is an explanatory diagram for describing the
clustering (classification) of the environment data
according to the second embodiment.
[0113]
In the present embodiment, the equipment data can be
classified into seven classes by using the On or Off signal
of the monitored facility.
[0114]
For example, a class 1 is a case where all of the
monitored facility A101, the monitored facility B102, and
the monitored facility C103 are turned off, a class 2 is a
case where the monitored facility A101 is turned on and the
monitored facility B102 and the monitored facility C103 are
turned off, a class 3 is a case where the monitored
facility B102 is turned on and the monitored facility A101
and the monitored facility C103 are turned off, a class 4
is a case where the monitored facility C103 is turned on
and the monitored facility A101 and the monitored facility
B102 are turned off, a class 5 is a case where the
monitored facility A101 and the monitored facility B102 are
turned on and the monitored facility C103 is turned off, a
class 6 is a case where the monitored facility B102 and the
monitored facility C103 are turned on and the monitored
facility A101 is turned off, and a class 7 is a case where
the monitored facility A101 and the monitored facility C103
are turned on and the monitored facility B102 is turned
off.
36
[0115]
That is, in the present embodiment, the environment
data is classified into the classes by using the On or Off
signals (operating data) of the plurality of monitored
facilities (the monitored facility A101, the monitored
facility B102, and the monitored facility C103), and the
environment data on which the noise removal processing is
to be executed is selected.
[0116]
As described above, it is possible to acquire the
environment data of each of the monitored facilities by
using the classes classified based on the On or Off signals
of the monitored facilities. The noise removal processing
can be executed.
[0117]
In the present embodiment, the class is classified by
using the On or Off signals of the pumps but there is a
high possibility that the sound information is changed by
the output in the pump. Thus, it is preferable that the
output signals (the current signals, the pressure signals,
and the flow rate signals) are classified together.
[0118]
In the present embodiment, the data storage unit 400
has such a clustering function (classification function),
and stores the environment data to which the label is given
in order to execute the noise removal processing and
facility specification processing for specifying the
37
monitored facility (this processing is the noise removal
processing in a broad sense).
[0119]
In the water plant (water purification plant), the
operation of the monitored facility may be changed
depending on the weather. Thus, it is preferable that
weather information (fine or rainy) is also classified
together. Accordingly, it is possible to execute the noise
removal processing and the facility specification
processing with higher accuracy.
[0120]
As described above, according to the present
embodiment, the environment data for detecting the state
changes of the monitored facilities can be saved
efficiently and selectively by using the operation data
without distinguishing between the steady state and the
non-steady state and with no prior scheduling.
[0121]
According to the present embodiment, it is possible
to more accurately execute the noise removal processing and
the facility specification processing by clustering
(classifying) the environment data on which the noise
removal processing or the facility specification processing
and storing the classified environment data, and it is
possible to more accurately detect the state change of the
monitored facility.
[0122]
38
In particular, in the present embodiment, since the
sound information is used as the environment data, spatial
information can be collectively acquired, and the state
changes of the plurality of monitored facilities can be
simultaneously detected.
[Third Embodiment]
[0123]
First, a monitoring control device 300 according to a
third embodiment will be described.
[0124]
FIG. 8 is an explanatory diagram for describing the
monitoring control device 300 according to the third
embodiment.
[0125]
In the present embodiment, an example in which a new
label is given to the environment data by using the
operation data other than the On or Off signal of the pump
and a cause of the state change of the pump is specified by
using whether or not the state of each of the environment
data and the operation data is changed when the state
change of the pump is detected will be described.
[0126]
As in the first embodiment, the monitoring control
device 300 described in the present embodiment has an
operation control function 310, a measurement data
reception function 320, an environment data preprocessing
function 330, a data transmission function 340, and an
39
alarm output function 350.
[0127]
In the present embodiment, in particular, in the
state change of the pump detected by the state change
detection means 500 is fed back into the environment data
preprocessing function 330, and the new label is given to
the environment data.
[0128]
Specifically, the environment data preprocessing
function 330 stores the preprocessed environment data of
which the data is analyzed by the state change detection
means 500. That is, the environment data preprocessing
function 330 stores the environment data after the
preprocessing is already executed and the state change
detection means 500 detects the state change of the pump.
[0129]
That is, for example, the state change detection
means 500 clusters (classifies) the operation data (the
result obtained by detecting the state change of the pump)
such as “the pressure signal of the pump exceeds an upper
or lower limit value”, “the current signal of the pump
exceeds a threshold value”.
[0130]
As in the first embodiment, when the state change of
the pump is detected, the state change detection means 500
outputs a detection signal of the state change to the alarm
output function 350.
40
[0131]
Meanwhile, the state change detection means 500
simultaneously outputs the detection signal of the state
change to the environment data preprocessing function 330.
[0132]
The environment data preprocessing function 330 give
a new label (Lc + numeral) to the environment data based on
the operation data when the state change is detected based
on the output signal of the input state change. That is,
the new label is given when the state change of the pump is
detected.
[0133]
As described above, the environment data to which the
new label is given overwritten and stored in the data
storage unit 400.
[0134]
Accordingly, the operation data in which the state
change of the pump is detected and the preprocessed
environment data can be stored appropriately and
selectively. When the state change of the pump is
detected, the cause of the state change of the pump can
also be specified by combining the preprocessed environment
data and the operation data.
[0135]
Next, a relationship between a failure occurring
around the pump and the operation data or the environment
data according to the third embodiment will be described.
41
FIG. 9 is an explanatory diagram for describing the
relationship between the failure occurring around the pump
and the operation data or the environment data according to
the third embodiment.
[0136]
For example, as illustrated in FIG. 9, the
relationship between the cause (failure occurring around
the pump) of the state change of the pump, the sound
pressure (environment data), and the operation data (the
pressure or the current) can be summarized by using the
environment data to which the new label is given, the
operation data at this time, and the cause of the state
change of the pump identified thereafter.
[0137]
When the state change detection means 500 detects the
state change of the pump, the sound pressure is changed,
the pressure is changed, and the current is not changed,
there is a high possibility that a “failure of a pump drive
unit” is a cause of the state change of the pump.
[0138]
As described above, the cause of the state change of
the pump can also be specified by combining the operation
data related to the pump and the preprocessed environment
data.
[0139]
Although it has been described in the present
42
embodiment that the pump is used as the monitored facility,
the monitored facility is not limited to such a pump. For
example, motors and other machineries can be similarly
used.
[0140]
Although it has been described that the pressure and
the current are used as the operation data, other operation
data such as the output and the flow rate may be used.
Although it has been described that the sound pressure of
the sound information is used as the environment data,
other sound information and other environment data such as
electromagnetic waves and odors may be used.
[0141]
According to the present embodiment, the new label
can be given to the environment data by using the operation
data other than the On or Off signal of the pump, and the
environment data and the operation data can be saved
appropriately and selectively. When the state change of
the pump is detected, the cause of the state change of the
pump can be specified by combining the operation data and
the environment data.
[Fourth Embodiment]
[0142]
First, a case where the label is given to the
environment data according to a fourth embodiment will be
described.
[0143]
43
FIG. 10 is an explanatory diagram for describing a
case where the label is given to the environment data
according to the fourth embodiment.
[0144]
In the present embodiment, a case where a label is
given to the environment data in consideration of an
inspection schedule of a worker (employee) in addition to
the On or Off signal of the pump will be described.
[0145]
In the present embodiment, the inspection schedule of
the worker is recorded in advance, and when the worker
works in a room where the target pump is installed, the
label may not be given to a time (a time for which the work
works in the room). That is, a label Lb8 and a label La8,
and a label La9 and a label Lb9 are given.
[0146]
Accordingly, it is possible to acquire only the
environment data necessary for detecting the state change
of the monitored facility by excluding the environment data
that does not contribute to the state change of the
monitored facility such as work sound emitted when the
worker works.
[0147]
The present invention is not limited to the
aforementioned embodiments, and includes various
modification examples. For example, the aforementioned
embodiments are described in detail in order to facilitate
44
easy understanding of the present invention, and the
present invention is not limited to necessarily include all
the components described in detail and specifically. Some
of the components of one embodiment can be replaced with
some of the components of another embodiment. It is also
possible to add the configuration requirements of another
embodiment to the configuration requirements of one
embodiment. It is also possible to add, delete, or replace
some of the components of another embodiment to, from, and
with some of the components of one embodiment.
[Description of Reference Numerals]
[0148]
10: plant state monitoring system
100: monitored facility
101: monitored facility A
102: monitored facility B
103: monitored facility C
200: environment data measurement device
300: monitoring control device
310: operation control function
320: measurement data reception function
330: environment data preprocessing function
340: data transmission function
350: alarm output function
400: data storage unit
500: state change detection means
45
[Name of Document] Claims
[Claim 1]
A plant state monitoring system, comprising:
an environment data measurement device that acquires
environment data of a monitored facility;
a monitoring control device that controls the
monitored facility, acquires operation data of the
monitored facility, acquires the environment data acquired
by the environment data measurement device, and executes
preprocessing of the environment data; and
a data storage unit that stores the environment data
acquired by the environment data measurement device and the
operation data acquired by the monitoring control device in
association with each other.
[Claim 2]
The plant state monitoring system according to claim
1, further comprising state change detection means for
detecting a state change of the monitored facility.
[Claim 3]
The plant state monitoring system according to claim
1, wherein the monitoring control device has an environment
data preprocessing function of executing the preprocessing
of the environment data, and gives a label to the
environment data together with a time by using the
operation data of the monitored facility and the
environment data of the monitored facility.
[Claim 4]
46
The plant state monitoring system according to claim
3, wherein the operation data is an ON or Off signal of the
monitored facility, and the environment data is sound
information of the monitored facility.
[Claim 5]
The plant state monitoring system according to claim
4, wherein the sound information is a frequency, a sound
pressure level, or a frequency component obtained by
acoustic analysis.
[Claim 6]
The plant state monitoring system according to claim
2, wherein the state change detection means uses an
adaptive resonance theory when the state change of the
monitored facility is detected.
[Claim 7]
The plant state monitoring system according to claim
3, wherein a label is given to the environment data earlier
and later than an input of the operation data.
[Claim 8]
The plant state monitoring system according to claim
7, wherein noise removal processing is executed by removing
the environment data to which the label is given later from
the environment data to which the label is given earlier or
removing the environment data to which the label is given
earlier from the environment data to which the label is
given later.
[Claim 9]
47
The plant state monitoring system according to claim
1, wherein the environment data is classified into classes
by using the operation data of the monitored facility, and
the environment data on which noise removal processing is
to be executed is selected.
[Claim 10]
The plant state monitoring system according to claim
1, wherein a cause of a state change of the monitored
facility is specified by combining the environment data and
the operation data.
[Claim 11]
A plant state monitoring method comprising:
acquiring environment data and operation data of a
monitored facility, and executing labeling processing for
giving a label to the environment data by using the
operation data; and
storing the environment data and the operation data
in association with each other.