FR3118181A1 - Method for characterizing an impulse electrical signal. - Google Patents

Abstract

Method for characterizing a pulsed electrical signal at the output of an electromechanical instrument, comprising steps consisting in:- emitting a plurality of pulsed electrical signals from the same electromechanical instrument,- for each pulsed electrical signal, recording the value of said pulsed electrical signal over time, - processing the recorded value of said pulsed electrical signal over time to deduce therefrom a set of parameters relating to said pulsed electrical signal,Characterized in that it further comprises: A step consisting in measuring and in recording the values as a function of time of the supply voltage at the origin of said pulsed electric signal,And in that the processing step comprises, simultaneously: - A frequency analysis step of said pulsed electric signal, and - A step of multiparameter analysis of said pulsed electrical signal.

Description

Method for characterizing an impulse electrical signal.

The present invention relates to the field of the characterization of a pulsed electrical signal.

Typically, an impulsive electrical signal is generated by an electromechanical instrument whose response is impulsive in nature or whose response is impulsive when subjected to an impulsive stimulus.

For brevity, electromechanical instrument means any type of instrument that generates current signals, which includes electrical and electromagnetic instruments.

The present invention is applicable to any electromechanical instrument which can be biased in alternating voltage, for example a sensor or a detector.

However, the invention can also be applied to characterize any pulsed electrical signal. Thus, problems related to electric shocks on electric motors, power lines, transformers, etc. can be characterized using the present invention.

A step of amplifying the electrical signals to be analyzed can be provided.

Alternatively, a step of attenuating the electrical signals to be analyzed can be provided.

Within the meaning of the present invention, the term “characterization” of a pulsed electrical signal is understood to mean: the location, in the instrument, of the phenomenon at the origin of said pulsed electrical signal, the characterization of the physics of the phenomenon at the origin of said pulsed electrical signal or the separation of two or more types of signals and therefore of different phenomena at the basis of the generation of the pulsed electrical signal.

As explained later, the characterization of a pulsed electrical signal from an electromechanical instrument makes it possible to determine its behavior.

According to a first of its objects, the invention relates to a method for characterizing a pulsed electrical signal at the output of an electromechanical instrument, comprising steps consisting of:
- emitting a plurality of pulsed electrical signals from the same electromechanical instrument,
- for each pulsed electrical signal, recording the value of said pulsed electrical signal over time,
- processing the recorded value of said pulsed electrical signal over time to deduce therefrom a set of parameters relating to said pulsed electrical signal.

It is essentially characterized in that it further comprises:
A step consisting in measuring and recording the values as a function of time of the supply voltage at the origin of said pulsed electrical signal,
And in that the processing step comprises, simultaneously,
A frequency analysis step of said pulsed electrical signal, and
A multiparameter analysis step of said pulsed electrical signal.

In one embodiment, the frequency analysis step consists in measuring the phase of the pulsed electrical signal when the amplitude of the latter is maximum with respect to the period of the sinusoidal voltage at the terminals of said electromechanical instrument,
The multiparametric analysis step consists of measuring and recording separately different characteristic values of said pulsed electrical signal, and
The processing step consists in measuring and recording separately different characteristic values of said pulsed electrical signal as a function of the phase shift between the phase of said pulsed electrical signal and the phase of the sinusoidal voltage at the terminals of said electromechanical instrument.

In one embodiment, the multiparametric analysis of said pulsed electrical signal comprises the measurement of at least one of the characteristics among: the amplitude, the partial load (Q _tail ), the crossing time (Drift _t ), the time rise of the signal (t _rise ), the fall time (t _fall ) and the total charge (Q _tot ) of said pulsed electrical signal.

In one embodiment, the electromechanical instrument is a fission chamber, the precedent further comprising a step consisting in detecting neutrons in a nuclear reactor thanks to said fission chamber.

In one embodiment, the method comprises a referencing step in which the processing step is implemented for a reference electromechanical instrument whose operation is nominal,
the method further comprising a step consisting in recording in a memory, as reference values, a set of values resulting from the frequency analysis step and a set of values resulting from the multiparametric analysis step.

In one embodiment, the method comprises a test step in which the processing step is implemented for a candidate electromechanical instrument whose operation is potentially non-nominal,
the method further comprising a step consisting in recording in a memory, as candidate values, a set of values resulting from the frequency analysis step and a set of values resulting from the multiparametric analysis step.

In one embodiment, the processing step is implemented until the counting statistics of the pulsed electrical signals are sufficient to satisfy a Poisson law.

In one embodiment, the method includes a step of modifying the value of an operating parameter.

In one embodiment, the method further comprises at least one of the steps from:
- a step consisting in amplifying said pulsed electrical signal; and
- A step of attenuating said pulsed electrical signal.

According to another of its objects, the invention relates to a computer program, comprising program code instructions for the execution of the steps of the method according to the invention, when said program is executed on a computer.

The present invention is not limited to electromechanical instruments, it can also be implemented to characterize the electrical signals related to electrical discharges on electric motors, power lines, transformers, etc. which concisely are also considered as "electromechanical instruments" within the meaning of the present invention.

Other characteristics and advantages of the present invention will appear more clearly on reading the following description given by way of illustrative and non-limiting example and made with reference to the appended figures.
Description of the drawings

illustrates an electrical diagram used for testing the method according to the invention,

illustrates the width at half maximum (FWHM) for one period (2π) of the power signal,

illustrates test results for a multiparametric and frequency characterization on 6 parameters, for nominal operation of a detector, and

illustrates test results for a multiparametric and frequency characterization on 6 parameters, for an operation of the detector of the , with a drop in supply voltage.
* Detailed description *

The present invention is applicable both to the field of detection and to that of the analysis of pulsed electrical signals, whether the signal is representative of a voltage or a current.

It allows, thanks to a coupling of frequency analysis and signal characterization, to achieve a better understanding of the causes of pulsed electrical signals, in particular those coming from detectors.

An embodiment of an electrical circuit used for testing the method according to the invention is illustrated in the .

As illustrated on the , an oscilloscope is connected to the terminals of an electromechanical instrument 10 whose behavior is to be characterized. In this case the electromechanical instrument 10 is a pressure sensor in an enclosure 20 whose pressure is adjustable.

A resistor R can be provided to protect the electromechanical instrument under test from excessive currents which could damage it.

A voltage probe is connected between the electromechanical instrument and the power supply.

A current probe is connected to the electromechanical instrument between the electromechanical instrument and the power supply or between the electromechanical instrument and ground, depending on the type of current probe used.

The signals from the two probes are then sent to an oscilloscope, then to a processor which processes them to create a two-dimensional image. Examples of two-dimensional images are shown in the and the .

The power supply is alternating, and has a maximum frequency preferably less than or equal to 1 MHz. All conventional waveforms can be implemented.

Provision is made to record the value of said pulsed electrical signal over time. The signal value is at least one of current intensity and voltage magnitude.

In this case, it is planned to measure an impulsive current at the output of the electromechanical instrument, and to record the values of the corresponding electrical intensity as a function of time.

It is also planned to measure and record the values of the supply voltage as a function of time at the input of the electromechanical instrument.

The parameters of the output current pulse, as well as the instant when the current pulse arrived with respect to the period (2π) of the supply signal are calculated as described later.

This process is repeated for a plurality of output current pulses.

For two output current pulses, the parameters of the output current pulse, as well as the instant when the current pulse arrived with respect to the period (2π) of the supply signal can be different.

It is thus possible to illustrate this phenomenon, as illustrated in the , generating an image that represents the spatial distribution of the output current pulses in a pattern, for one period of the power supply signal.

In this case, the image is obtained by stacking several output current pulses.

A two-dimensional histogram or three-dimensional graphic is thus obtained, in this case using a scale of gray levels, or preferably, using a color scale.

The illustrates test results for one period (2π) of the power signal. Each point corresponds to the width at half maximum (FWHM) versus phase, in this case in degrees, for a given output pulse. A color code, in this case graduated between 0 and 6, makes it possible to count the number of pulses having the same characteristics.

* Example *

By way of illustrative, non-limiting example, the present invention is applied to a fission chamber, which is an ionization detector on which is placed a layer of fissile material, used to detect neutrons in a nuclear reactor and which generally comprises several electrodes. One or more electrodes are biased at a nominal voltage and the inter-electrode space is filled with a noble gas. When a neutron passes through the fission chamber, depending on its interaction cross section, it has a strong chance of interacting with the fissile material deposited on the electrodes generating a fission reaction. When this happens, a current pulse is then created and then measured. The characteristics of the pulse coming from a fission chamber depend on the geometry of the chamber, the type and the pressure of the gas inside, as well as the perfect functioning of the electric cables which carry the aforementioned pulse.

A fission chamber therefore generates a pulsed electrical signal when it is subjected to radioactivity. One of the modes of operation of a fission chamber is to use its impulse response to count the impulses and deduce the operating parameters of the reactor.

More generally, a sinusoidal voltage is applied to the terminals of an electromechanical instrument and its current impulse response is observed. We can then implement a frequency analysis. The phase shift between the impulse response and the phase of the voltage makes it possible to characterize the behavior of the electromechanical instrument.

As such, the illustrates the distribution of the half-maximum widths of the impulse responses of a signal with respect to the phase of the input voltage. Each dot represents a pulse, and the color or gray level of the dot represents the number of pulses exhibiting that characteristic.

We can thus extract information such as the location of these points, that is to say their phase shift and their dispersion.

The multiparametric analysis of the pulsed electrical signal consists in measuring and recording separately different characteristic values of said signal, for example its amplitude, its total area, its time decay, etc.

Such a multiparametric and frequency analysis is illustrated for example on the which represents 6 analysis criteria, or parameters or even characteristics, of a pulsed signal as a function of the phase of the voltage signal: in this case the amplitude, the partial load (Q _tail ), the crossing time ( Drift _t ), signal rise time (t _rise ), fall time (t _fall ) and total load (Q _tot ).

Thanks to this simultaneous combination of frequency analysis and multiparametric analysis, the invention makes it possible to characterize the operation of a detector, in particular its aging.

For example in the field of a nuclear reactor, a fission chamber allows the monitoring of its power. When a fission chamber is subjected to high temperatures or high pressures, or during its aging, it undergoes a phenomenon of partial discharges which generate pulses inside the detector which are not the result of interactions. The pulse signals thus generated are therefore false positives, the fission chamber does not necessarily count neutrons.

Thanks to the present invention, thanks to the multiparametric and frequency analysis of the dispersion of the pulsed signals, it is possible to detect that the pulsed signals come from partial discharges and to characterize for example the aging of the fission chamber, for example linked to a drift due to pressure or temperature.

It is thus possible to identify that the detector is not operating at nominal speed, which is an advantage of the present invention. In addition, as described later, it can also make it possible to characterize the deviation from nominal operation.

The present invention is not limited to the field of radiation detectors.

It can for example be implemented in the field of electrical discharges. In this field, it is useful to characterize the operation of an electromechanical instrument which generates electrical discharges within an electrical connection thereof.

The present invention makes it possible to characterize the nominal operation of an electromechanical instrument.

On the , the present invention is implemented during nominal operation of said electromechanical instrument, the supply voltage at the terminals of said electrical connection being stable.

On the , the present invention is implemented during a drop in the supply voltage across the terminals of said electrical connection.

By comparing the and the , it is clearly observed that the distribution of the pulse signals is different.

It is thus possible to record these distributions in a memory, which thus become reference signals.

Preferably, recording continues until the pulse signal count statistics are sufficient to satisfy a Poisson law.

Thanks to this recording, one can then implement the invention again on a candidate electromechanical instrument, which can be the same electromechanical instrument at a later time or another electromechanical instrument of the same type, and compare the results to the reference signals. . Depending on the result of the comparison, it is thus possible to characterize, for example, a voltage drop within the candidate electromechanical instrument.

On this basis, it is possible to provide a referencing step consisting in recording in a memory, for example a database, the reference pulse signals of an electromechanical instrument according to the invention in nominal operation.

It is then possible to modify the value of a single operating parameter at a time, for example the supply voltage, the temperature, the pressure, etc., and to record the corresponding pulse signals, in association with the nature of said operating parameter. .

Provision can also be made to record the value of said operating parameter, in absolute value, for example the temperature T = 300 K, or in relative value, for example the temperature T = nominal temperature + 50 K, etc.

Provision can be made to implement this step for a plurality of values of an operating parameter, so as to obtain a distribution of respective pulse signals for a respective value of a respective operating parameter. For example a first distribution of pulse signals for a first temperature value, a second distribution of pulse signals for a second temperature value, etc.

It thus becomes possible subsequently not only to characterize the phenomenon at the origin of the non-nominal operation, that is to say to qualify it, for example a drop in voltage, but also to quantify it, for example a voltage drop of 10V or a voltage drop of 25V, etc.

For example, the nature of the operating parameter (supply voltage, temperature, pressure, etc.) and possibly its value(s) are recorded in specific fields of a database or in the form of metadata.

In the field of signal characterization, it is known to extract a set of signal parameters.

In the field of frequency analysis, it is known to analyze the response of a system by the position of its impulse response with respect to the sinusoid of the input signal to which it is subjected.

The present invention advantageously aims to combine these two teachings to extract from the frequency analysis of the pulsed signals, a set of multiple parameters which make it possible to characterize a behavior of the system from which said pulsed signals originate.

The characterization of signals according to the invention is the synchronization of a multiparametric and frequency analysis of the signals. It already makes it possible to improve the characterization of the system from which said pulse signals originate.

In addition, by combining multiparametric analysis and frequency analysis of signals applied simultaneously, it is possible, thanks to the invention, to characterize a behavior of the system from which said pulsed signals originate according to criteria that are impossible to observe. with one of these two methods taken separately.

For example, the present invention allows a better characterization, a better understanding of the system from which the pulsed signals come, or in the case of a fission chamber, a better discrimination of the particles at the origin of the pulsed signals.

A single parameter may not be sufficient to characterize non-nominal behavior. For example on the , the values of the fall time t _fall for non-nominal operation (voltage drop) are difficult to distinguish from the values of the fall time t _fall of the nominal operation illustrated on the .

It is therefore useful to use several parameters.

Thanks to the synchronization of the multiparametric and frequency analysis of the impulse signals according to the present invention, it is then possible to compare the impulse response of a candidate electromechanical instrument on another parameter and thus to be able to nevertheless characterize a nominal or non-nominal behavior. of an electromechanical instrument.

Between nominal operation and non-nominal operation of an electromechanical instrument, there may or may not be overlap points. The impulse response of nominal operation and non-nominal operation is specific to each electromechanical instrument.

In the field of fission chambers, at present, the detectors are taken out of the reactor, characterized outside the reactor, for example in the laboratory.

Thanks to the present invention, it is possible to characterize a fission chamber during the operation of a reactor.

For example, if a detector emits suspect output signals, for example different from the other detectors, it is thus possible to characterize the current behavior of said suspect detector without stopping the reactor, by comparison with the characterization of its nominal operation recorded in a memory. .

Advantageously, the same electromechanical instruments have substantially the same nominal operating characteristics. For example, fission chambers with the same geometries also have globally the same nominal operating characteristics.

It is therefore not necessary to characterize all electromechanical instruments individually.

It is possible to characterize a single electromechanical instrument for a set of electromechanical instruments of the same category.

Claims (10)

Method for characterizing a pulsed electrical signal at the output of an electromechanical instrument, comprising steps consisting of:
- emitting a plurality of pulsed electrical signals from the same electromechanical instrument,
- for each pulsed electrical signal, recording the value of said pulsed electrical signal over time,
- processing the recorded value of said pulsed electrical signal over time to deduce therefrom a set of parameters relating to said pulsed electrical signal,
Characterized in that it further comprises:

• A step consisting in measuring and recording the values as a function of time of the supply voltage at the origin of said pulsed electrical signal,

And in that the processing step includes, simultaneously:

• A frequency analysis step of said pulsed electrical signal, and
• A multiparameter analysis step of said pulsed electrical signal.

A method according to claim 1, wherein:

• The frequency analysis step consists in measuring the phase of the pulsed electrical signal when the amplitude of the latter is maximum with respect to the period of the sinusoidal voltage at the terminals of said electromechanical instrument,
• The multiparametric analysis step consists of measuring and recording separately different characteristic values of said pulsed electrical signal, and
• The processing step consists in measuring and recording separately different characteristic values of said pulsed electrical signal as a function of the phase shift between the phase of said pulsed electrical signal and the phase of the sinusoidal voltage at the terminals of said electromechanical instrument.

Method according to any one of the preceding claims, in which the multiparametric analysis of the said pulsed electrical signal comprises the measurement of at least one of the characteristics among: the amplitude, the partial load (Q _tail ), the crossing time ( Drift _t ), the rise time of the signal (t _rise ), the fall time (t _fall ) and the total charge (Q _tot ) of said pulsed electrical signal. A method according to any preceding claim, wherein the electromechanical instrument is a fission chamber, the method further comprising a step of detecting neutrons in a nuclear reactor using said fission chamber. Method according to any one of the preceding claims, comprising a referencing step in which the processing step is implemented for a reference electromechanical instrument whose operation is nominal,
the method further comprising a step consisting in recording in a memory, as reference values, a set of values resulting from the frequency analysis step and a set of values resulting from the multiparametric analysis step. Method according to any one of the preceding claims, comprising a test step in which the processing step is carried out for a candidate electromechanical instrument whose operation is potentially non-nominal,
the method further comprising a step consisting in recording in a memory, as candidate values, a set of values resulting from the frequency analysis step and a set of values resulting from the multiparametric analysis step. Method according to any one of the preceding claims, comprising a step consisting in recording in a memory the distribution of the pulsed electrical signals and establishing a counting statistics of the pulsed electrical signals; wherein the processing step is carried out until the counting statistics of the pulsed electrical signals are sufficient to satisfy a Poisson law. A method according to any of claims 4 to 7, wherein the fission chamber is used to detect neutrons in a nuclear reactor; in which one of the modes of operation of the fission chamber is to use the impulse response of the latter to count the impulses and deduce therefrom the operating parameters of the reactor; the method comprising a step of modifying the value of an operating parameter. A method according to any preceding claim, further comprising at least one of:
- a step consisting in amplifying said pulsed electrical signal; and
- A step of attenuating said pulsed electrical signal. A computer program, comprising program code instructions for carrying out the steps of the method according to any preceding claim, when said program is executed on a computer.