TECHNICAL TESTING METHOD
Background of the invention
The present invention relates to a technical test
5 method for testing a device and including operating the
device in at least one operating stage that corresponds
to a stable value of at least one operating setpoint for
the device and/or for a test bench for testing the
device.
10 In the field of technical testing, it is common
practice to test devices at a plurality of different
operating rates that are to be encountered while the
device is in use, with this being for the purpose of
characterizing the entire operating envelope of the
15 device as completely as possible. Typically, during
testing, th;is is done by following a sequence comprising
a plurality of operating stages, with each operating
stage being maintained for a predetermined duration that
is considered as being sufficient for guaranteeing the
20 quality of information about operating parameters of the
device as obtained over the duration of this stage.
Nevertheless, running such a sequence of operating
stages of predetermined duration can result in tests
presenting an overall duration that is considerable.
25 Unfortunately, and mainly for economic reasons, it is
desirable to limit the overall duration of testing. In
addition, adjusting the individual duration of each
operating stage would make it possible to test a larger
number of operating setpoint values, thereby enabling the
30 operating envelope of the device subjected to testing to
be characterized more completely.
35
Object and summary of the invention
The technical test method described in the present
disclosure seeks to remedy those drawbacks. In
particular, this disclosure seeks to provide a technical
test method that makes it possible to reduce the overall
2
duration of a technical test and/or to increase the
number of stages in a sequence of operating stages for
the device under test, while maintaining the validity of
the test.
5 In at least one implementation, this object is
achieved by the fact that, during the technical test,
1vhich comprises at least one operating stage
corresponding to a stable value of at least one operating
setpoint for the device and/or for a test bench for
10 testing the device, said operating stage is finalized
before a maximum duration threshold if a criterion
associated \·lith a set of physical parameters picked up
during the operating stage is satisfied and if a
confidence level associated with said set of physical
15 parameters reaches at least a predetermined threshold.
This second condition makes it possible to distinguish
between situations in which the criterion is sufficiently
reliable for it to be opportune to finalize the operating
stage, and situations in which. the criterion is not
20 reliable, and which 1vould thus not enable the operating
stage to be shortened, even if the criterion is
satisfied. In particular, an additional criterion for
finalizing said operating stage before a maximum duration
threshold may be that said criterion is satisfied and
25 that said confidence level has reached at least said
predetermined threshold, for at least some predetermined
minimum duration, in order to avoid some transient
fluctuation in.a physical parameter of said set of
parameters triggering premature interruption of the
30 operating stage.
In this context, the term ''set of physical
parameters" should be understood broadly, and can thus
comprise a single physical parameter. Nevertheless, said
set of physical parameters may comprise a plurality of
35 physical parameters each associated 1-1ith a respective
confidence level, the confidence level associated 1-1ith
the set of physical parameters as a 1-1hole being a
3
function of the confidence levels associated with said
plurality of physical parameters. Thus, the reliability
of each of the physical parameters may be weighted
depending on the importance of the physical parameter in
5 calculating the confidence level that is associated with
the plurality of physical parameters and that authorizes
the transition to the following operating stage. By way
of example, said function may comprise the product of
multiplying together the confidence levels associated
10 with two physical parameters of said· plurality, and/or
subtracting the product of multiplying together the
confidence levels associated with two physical parameters
of said plurality from the sum of the same two confidence
levels. In fuzzy logic, the product of two truth values
15 corresponds to the probabilistic t-norm operator, whereas
subtracting the product of two truth values from the sum
of the same two truth values corresponds to a
probabilistic t-conorm operator.
In order to match the confidence level corresponding
20 to each physical parameter to the available information
about the reliability of that physical parameter, the
confidence level may in particular be predetermined, or
calculated as a function of a noise level, and/or as a
function of an asymmetric uncertainty coefficient in a
25 signal corresponding to the associated physical parameter
during a moving time window, and/or a difference between
a value of the associated physical parameter and a
predetermined threshold.
In order to be able to apply the principles of fuzzy
30 logic to processing confidence levels, each of said
confidence levels may have a value lying in the range 0
to 1.
In order to be able to test the device at a
plurality of different operating rates, the method may
35 comprise a sequence of a plurality of different operating
stages, each corresponding to a stable value of at least
one operating setpoint for a device subjected to testing
4
and/or for a test bench for testing the device. The
order of the operating stages in said sequence may be
established on the basis of at least one priority
assigned to each operating stage, and of values for the
5 at least one operating setpoint corresponding to the
plurality of operating stages. The order of the
operating stages may be modified, on the basis of
predefined criteria, depending on how the device being
subjected to the technical test responds.
10 The device subjected to the technical test may in
particular be an engine, in particular a liquidpropellant
rocket-engine, and more specifically a liquidpropellant
rocket-engine having a turbopump feed system.
The present invention also provides an electronic
15 control unit having at least one data output for
transmitting at least one operating setpoint to a device
and/or to a test bench for testing said device, the unit
being configured to control a technical test of the
device in application of the above method. This
20 configuration may be a physical configuration of at least
one electronic circuit of the electronic control unit, or
it may be implemented in a programmable electronic
control unit by means of software, i.e. a set of
instructions executable by a computer system for
25 performing a technical test method. Such a set of
instructions may be contained in a data medium. The term
"data medium" designates any data storage device capable
of being read by a computer system. Such a data medium
may in particular be a magnetic data storage device, such
30 as a magnetic disk or tape, or an optical data storage
device such as an optical disk, or an electronic data
storage device, such as a volatile or non-volatile
electronic memory.
35 Brief description of the drawings
The invention can be \vell understood and its
advantages appear better on reading the follmving
5
detailed description of an implementation shown by way of
non-limiting example. The description refers to the
accompanying drawings, in which:
· Figure 1 is a diagram showing a liquid-propellant
5 rocket-engine that is fed by turbopumps on a test bench
including an electronic control unit in an embodiment of
the present invention;
· Figure 2 is a diagram showing the six main
functions of a technical test method in an implementation
10 of the invention;
15
Figure 3A is a graph showing the level of
confidence attributed to a signal as a function of the
dispersion of noise in the signal in a moving time
w indmv; and
· Figure 3B is a graph showing the level of
confidence attributed to a signal as a function of its
asymmetric uncertainty coefficient in a moving time
1vindov1.
20 Detailed description of the invention
Figure 1 shows a liquid-propellant rocket-engine 1
that is fed by turbopumps, the engine being installed on
a test bench 2 in which it is connected to an electronic
control unit 3 for performing tests using a profile
25 comprising a sequence of a plurality of stages of
operation.
In the embodiment shown, the rocket engine 1 "is a
rocket engine of the "expansion cycle" type, in which the
"turbopumps 4 and 5 are actuated by one of the propellants
30 after passing through a regenerative heat exchanger 6
adjacent to the walls of the propulsion chamber 7 of the
rocket engine 1.
bet1·1een the tanks
Feed valves 8 and 9 are interposed
10 and 11 containing the propellants
and the corresponding turbopumps 4 and 5, and bypass
35 valves 12 and 13 enable these turbopumps 4 and 5 to be
bypassed at least in part by the propellant heated by the
heat exchanger 6. Nevertheless, the invention is not
6
·limited in any way to testing such rocket engines, and it
may equally well be applied to testing other types of
engine and indeed other types of device.
In the embodiment shmm, the operation of the rocket
5 engine 1 can be controlled by means of the feed valves 8
and 9 and the bypass valves 12 and 13. Each of these
valves is connected for this purpose to the electronic
control unit 3 in order to receive operating setpoints.
The test bench 2 also has sensors, such as for example
10 temperature and pressure sensors 14 and 15 in the
propulsion chamber 7, and thrust and vibration sensors 16
and 17 in the supports of the rocket engine 1. These
senso~s 14, 15, 16, 17 are also connected to the
electronic control unit 3 in order to transmit operating
15 parameters of the rocket engine 1 thereto. This set of
operating parameters X may include a first parameter A, a
second parameter B, and so on.
In the test bench 2, the rocket engine 1 is to be
subjected to technical tests comprising a sequence of
20 operating stages PFn in order to evaluate the operating
parameters X of the rocket engine 1 as picked up during
each of these operating stages. Each operating stage PFn
in this sequence corresponds to a set of stable setpoint
values for the operation of the rocket engine 1 and seeks
25 to reproduce operating points that are pertinent for
normal utilization of the rocket engine 1.
Each operating stage needs to be of a duration that
is sufficie~t to collect values that are representative
of operating parameters X of the rocket engine 1 during
30 such a test. Simultaneously, an excessive duration for
operating stages, and thus for the test, presents
drawbacks, in particular in
therefore necessary to find
of the stages.
terms of cost. It is
a compromise for the duration
35 Figure 2 shows the six main steps of a technical
test method that can be performed with the system shmm
in Figure 1. In a first function F1 of the method, a
7
table comprising data about each operating stage PFn is
stored in a memory (data storage device or random access
memory (RAM)) of the electronic control unit 3. For each
operating stage PFn, this table may comprise in
5 particular: an identifier of the operating stage PFn, a
first parameter para1,n giving a priority assigned to the
stage PFn; operating setpoint values and possibly also
monitoring thresholds for the rocket engine 1 and/or the
test bench 2; a maximum duration tmax,n and possibly a
10 minimum duration tmin,n for the stage; at least one end-ofstage
criterion, each criterion being associated with a
set of physical operating parameters X picked up during
the operating stage; a confidence level associated with
each set of operating parameters X associated with each
15 end-of-stage criterion; and logic and/or mathematical
operators to be applied to said criteria, physical
parameters, and/or confidence levels. By 1vay of example,
the parameter para1 ,n giving the priority assigned to each
stage PFn may have a value selected from the values 1, 2,
20 and 3, the value 1 corresponding to the highest priority
and the value 3 to the lowest priority. By way of
example, the end-of-stage criteria may be criteria
concerning the stability of said operating parameters X
as sampled during a moving time window.
25 In a second function F2, the electronic control unit
3 orders the operating stages from the table established
in the first function Fl following the priority given to
each stage, but while also taking account of the
proximity of the values for operating setpoints bet1·1een
30 successive stages. In certain particular circumstances,
t1w or more operating stages may be associated in order
to guarantee that they are executed in a previously
imposed order. This second function F2 is executed at
the beginning of the technical test, and is then repeated
35 before the end of each operating stage PFn in order to
update the sequence of the stages that remain. In this
implementation, the electronic control unit can also take
8
account of the time remaining until the fuel available in
the tanks 10 and 11 has been spent.
Thus, by 1·1ay of example, each operating stage PFn may
be associated with said first parameter para1 , 0 , with a
5 second parameter para2,n corresponding to its proximity to
a current stage, and with a third parameter para3,n
corresponding to the difference between the maximum
duration tmax,n assigned thereto and the maximum operating
duration tcap,n that can be maintained at the rate
10, associated with this operating stage PFn and with the
quantities of the propellants still available in the
tanks 10, 11.
In this example, the value of the second parameter
para2,n for each stage that remains to be selected for
15 this sorting may be calculated using the following
formula, by way of example:
para2,n = ~a1• (p0 - Pc)
2 + a 2 • (RM0 - RMc)
2
where Pn and RMn are respectively the combustion pressure
and the propellant mixing ratio during this operating
20 stage PF0 , and Pc and RMc are respectively the present
pressure and the propellant mixing ratio in the
propulsion chamber 7, 1vhereas a 1 [ALPHA_l] and a2
[ALPHA_2] are coefficients for weighting these physical
parameters characterizing the stage.
25 The value of the third parameter para3,n may be
calculated for each operating stage PFn that remains by
subtracting the value of the maximum duration tmax,n
assigned thereto from the maximum operating duration tcap,n
corresponding to this rate of operation and to the
30 remaining propellant capacity.
If for at least one of the remaining operating
stages PFn the value of the third parameter para3,n is
negative, thus indicating that the stage in question
cannot be maintained for the maximum duration tmax,n
35 assigned thereto without running out of propellant, then
the remaining stage(s) satisfying this condition may be
•_:_j
~
9
selected and classified in increasing order of the
weighted sum Kn of the first parameter para1,n and of the
second parameter para2,n using the following formula:
Kn = lh.paral,n + P2.para2,n
5 where p, [BETA_l] and P2 [BETA_2] are coefficients
respectively for weighting said first and second
parameters para1,n and para2,n.
In contrast, if none of the remaining stages
satisfies this condition, then all of ·the remaining
10 stages are selected and classified by increasing order of
the \veigh ted sum K' n of the values not only of the first
parameter para1,n and of the second parameter para2,n, but
also of the third parameter para3,n, using the follO\ving
formula:
15 K'n = p,.para1,n + P2.para2,n + p3.para3,n
where P3 [BETA_3] is a weighting coefficient for said
third parameter para3 ,n.
In a third function F3, the electronic control unit
3 generates setpoints for transmitting to the rocket
20 engine 1 and/or to the test bench 2, together with the
corresponding monitoring thresholds, on the basis of data
stored in the table for the operating stage PFn in the
first position (i.e. presenting the lowest value for the
weighted sum Kn or K'n) in the most recent classification
25 established by the second function F2. Thereafter, in
the fourth function F4, these setpoints are applied by
the electronic control unit 3 in order to control the
operation,qf the rocket engine 1 and/or of the test bench
2.
30 Concurrently with the fourth function F4, the
electronic control unit 3 performs a fifth function F5 of
picking up and processing physical parameters X
associated with the current operating stage PFn, in
particular by means of the sensors 14 to 17. The values
35 of these physical parameters may be picked up with
sampling at high frequency or at lm1 frequency depending
on the parameter and on its frequency range that is to be
10
analyzed. For example, for signals to be analyzed in a
range lower than 25 hertz (Hz), it is possible to apply
low frequency sampling at approximately 100 points per
second (pt/s), 1vhereas for signals that need to be
5 analyzed over a range that may be substantially higher
than 25 Hz, and that may even reach 5000 Hz, it is
possible to apply sampling at a high frequency of about
25,000 pt/s.
In this fifth function F5, a confidence level CLx in
10 the range 0 to 1 may be associated with each signal
corresponding to one or more physical parameters X, as a
function in particular of an estimated reliability of the
sensor and/or of algorithms used for the processing of
each signal. The value of each of these confidence
15 levels CLx may be predetermined, or it may be calculated
in real time as a function of a noise level and/or a bias
level in the signal corresponding to the associated
physical parameter, and/or as a function of the
difference between the signal and a predetermined
20 threshold. Thus, by 1vay of example, it is possible to
allocate said confidence level CLx to each signal sampled
at lov1 frequency as a function of the level of noise in
the signal, where noise is measured by means of the
dispersion cr [SIGMA] of the signal in a moving time
25 wind01-1, and to allocate said confidence level CLx to each
signal sampled at high frequency as a function of the
asymmetric uncertainty coefficient y1 [GAMMA_l] of the
signal in a moving time wind01v.
The dispersion cr [SIGMA] of a signal and the
30 confidence level CLx allocated to the signal may comply
with the relationship sh01m in Figure 3A, for example.
In this example, a high confidence level CLx is not
allocated to the signal if its dispersion cr [SIGMA] is
too great, indicating a high level of noise, or on the
35 contrary if it is too small, indicating that the signal
is too constant (dead signal).
11
The asymmetric uncertainty coefficient y1 [GAMMA_l)
of a signal and the confidence level CLx allocated to the
signal may, for their part, comply with the function
shown in Figure 3B, for example. In this example, a high
5 confidence level CLx is not allocated to this signal if
the absolute value of its asymmetric uncertainty
coefficient Y1 [GAMMA_l) is too high.
It is also possible to associate a confidence level
CLset to a set of physical parameters X, with this
10 confidence level CLset being calculated by the electronic
control unit 3 on the basis of the values of the
confidence level CLx individually associated 1-1ith the
physical parameters X of the set. This calculation may
be performed in particular by applying fuzzy logic
15 operators to the values of the confidence levels CLx
individually associated with the physical parameters X of
the set. Among the operators that may be used, there are
in particular probabilistic operators, and in particular
the probabilistic t-norm and t-conorm operators. The
20 first corresponds to the product of two terms of the
operation, while the second corresponds to subtracting
the product of the two terms of the operation from the
sum of the two terms of the operation.
Thus, by way of example, by applying the t-norm
25 operator, the confidence levels CLA, CL", and CLc of three
signals corresponding respectively to physical parameters
A, B, and C contributing to a single end-of-stage
criterion are multiplied together to obtain the
confidence level CLset associated with this set of
30 physical parameters, and thus 1·lith a corresponding endof-
stage criterion.
Finally, in a sixth function F6, the electronic
control unit 3 verifies lvhether the at least one end-ofstage
criterion is satisfied and whether at least one
35 confidence level associated with this criterion has
reached a minimum threshold. If the at least one
criterion is satisfied and the confidence level has
12
reached the minimum threshold, possibly for at least some
minimum length of time, the current stage can be
finalized and the electronic control unit 3 can return to
the second function F2 in which the table is re-arranged
5 after eliminating the stage that has been finalized,
after which the third function F3 generates the operating
setpoints that correspond to the following stage.
Finalizing the current stage may also depend on a minimum
duration for that stage, so that the operating stage is
10 not finalized too quickly.
In contrast, if the at least one end-of-stage
criterion is not satisfied and/or if the at least one
confidence level associated with this criterion has not
reached its minimum threshold, the stage may be continued
15 until a predetermined maximum duration is reached.
20
Although the present invention is described with
reference to a specific implementation, it is clear that
various modifications and changes may be made on these
implementations without going beyond the general ambit of
the invention as defined by the claims. In addition,
individual characteristics of the various implementations
ment'ioned may be combined in additional implementations.Consequently, the description and the drawings should be considered in a sense that is illustrative rather than restrictive.
13

CLAIMS
1. A technical test method for testing a device, the method comprising at least one operating stage
corresponding to a stable value of at least one operating setpoint for the device and/or for a test bench for
testing the device, wherein said operating stage is
finalized before a maximum duration threshold if a
criterion associated with a set of physical parameters
picked up during the operating stage is satisfied and if
10 a confidence level associated with said set of physical
parameters reaches at least a predetermined threshold.
2. A technical test method according to claim 1, 1·1herein
said operating stage is finalized prior to a maximum
15 duration threshold if said criterion is satisfied and if
said confidence level reaches at least said predetermined
threshold for at least a predetermined minimum duration.
3. A technical test method according to claim 1 or claim
20 2, wherein said set of physical parameters comprises a
plurality of physical parameters each associated with a
respective confidence level, the confidence level
associated with the set of physical parameters as a whole
being a function of the confidence levels associated 1vi th
25 said plurality of physical parameters.
4. A technical test method according to claim 3, wherein
said function comprises the product of multiplying
together the confidence levels associated with two
30 physical parameters of said plurality.
5. A technical test method according to claim 3 or claim
4, wherein said function comprises subtracting the
product of multiplying together the confidence levels
35 associated with two physical parameters of said plurality
from the sum of the same two confidence levels.
14
6. A technical test method according to any one of claims
3 to 5, wherein at least one of said confidence levels is
predetermined.
5 7. A technical test method according to any one of claims
3 to 6, wherein at least one confidence level associated
with a physical parameter is calculated as a function of
a noise level in a signal corresponding to the associated
physical parameter.
10
8. A technical test method according to any one of claims
3 to 7, wherein at least one confidence level associated
with a physical parameter is calculated as a function of
an asymmetric uncertainty coefficient of a signal
15 corresponding to the associated physical parameter.
9. A technical test method according to any one of claims
3 to 8, wherein at least one confidence level associated
with a physical parameter is calculated as a function of
20 a difference between a value of the associated physical
parameter and a predetermined threshold.
10. A technical test method according to any one of
claims 3 to 9, wherein each of said confidence levels has
25 a value lying in the range 0 to 1.
11. A technical test method according to any preceding
claim, comprising a sequence of a plurality of different
operating stages, each corresponding to a stable value of
30 at least one operating setpoint for a device subjected to
testing and/or a test bench for the device.
12. A technical test method according to claim 11,
wherein the order of the operating stages in said
35 sequence is established on the basis of at least one
priority assigned to each operating stage, and of values
15
for the at least one operating setpoint corresponding to
the plurality of operating stages.
13. A technical test method according to any preceding
5 claim, wherein the device is an engine.
10
14. A technical test method according to claim 13,
wherein said engine is a liquid-propellantlrocket-engine
( 1) .
15. A technical test method according to claim 14,
wherein said liquid-propellant rocket-engine (1) includes
a turbopump feed system.
15 16. An electronic control unit (3) having at least one
data output for transmitting at least one operating
setpoint to a device and/or to a test bench (2) for
testing said device, the unit being configured to control
a technical test of the device in application of the
20 method according to any one of claims 1 to 15.
25
17. A data medium containing a set of instructions
executable by a computer system to perform a technical
test method according to any one of claims 1 to 15.