CN115329810A - Traction motor health diagnosis method and system - Google Patents
Traction motor health diagnosis method and system Download PDFInfo
- Publication number
- CN115329810A CN115329810A CN202210943613.3A CN202210943613A CN115329810A CN 115329810 A CN115329810 A CN 115329810A CN 202210943613 A CN202210943613 A CN 202210943613A CN 115329810 A CN115329810 A CN 115329810A
- Authority
- CN
- China
- Prior art keywords
- fault
- signals
- traction motor
- equal
- angle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/34—Testing dynamo-electric machines
- G01R31/343—Testing dynamo-electric machines in operation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Abstract
The invention provides a traction motor health diagnosis method and a system, wherein the method comprises the following steps: the method comprises the steps that a plurality of existing electric signals of a traction motor are obtained from a traction control unit of a converter, and equal-angle resampling and data preprocessing are respectively carried out on the existing electric signals to obtain a plurality of equal-angle dimensionless stable signals irrelevant to the rotating speed, the working condition and the load; respectively demodulating a plurality of the equal-angle dimensionless steady signals and extracting a plurality of fault characteristic indexes by adopting a signal time-frequency analysis method; taking a plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis, and outputting a fault mode of the traction motor and a severity degree corresponding to the fault mode; and carrying out health management on the traction motor according to a plurality of fault characteristic indexes. The method can effectively eliminate the diagnosis errors caused by transient processes such as PWM power supply harmonic waves, variable rotating speed and variable load, and the like, and improves the accuracy of non-stable weak signal characteristic extraction and fault diagnosis.
Description
Technical Field
The invention belongs to the technical field of fault prediction and health management of a rail transit train traction system, and particularly relates to a traction motor health diagnosis method and system.
Background
In order to improve the safety guarantee, the application efficiency and the economy of train vehicles, improve the failure analysis efficiency, reduce the preparation time and construct a safe, efficient, green and economic intelligent train system, the trend is great. Therefore, the method researches the technical scheme of health diagnosis and prediction of key parts of the rail transit train including the traction motor, dynamically monitors the health state of the traction motor to realize 'state repair' of the traction motor, and has important significance for improving the running safety and intelligent operation and maintenance of the train.
In the prior art, a vibration method, a temperature method, an acoustic signal method, an electric signal method and the like are mainly adopted for detecting the faults of the traction motor. The electric signal analysis method can diagnose electric and mechanical faults simultaneously, does not need to add monitoring equipment such as a sensor in field application, and has the characteristics of high coverage rate of diagnostic points and non-invasive type, so that the electric signal analysis method is gradually and widely concerned and intensively researched. Aiming at the health diagnosis of the traction motor, an electric signal analysis method has the technical difficulties that: the essence of the electrical signal analysis is to extract well-known characteristic signal components resulting from specific traction motor faults from the converter existing electrical signals, these fault indicators depending only on the type of fault, the motor speed and the size of the motor mechanical structure. However, these weak characteristic signal components are often buried in the environmental noise due to the strong background noise such as torque ripple and power supply harmonics caused by the traction assistance. In addition, due to the influences of actual service conditions such as frequent working condition conversion, strong time variation of rotating speed, load disturbance and the like of the traction motor, the acquired signals present non-stationarity, which is specifically reflected in that the amplitude and the frequency of the characteristic components change along with the time variation, and thus great challenges are brought to the extraction and accurate identification of fault characteristic indexes.
Disclosure of Invention
The invention provides a traction motor health diagnosis method and system, which aim to solve the problem that fault characteristic indexes are difficult to accurately extract for fault diagnosis and early warning in the prior art.
Based on the above purpose, an embodiment of the present invention provides a traction motor health diagnosis method, including: the method comprises the steps that a plurality of existing electric signals of a traction motor are obtained from a traction control unit of a converter, and equal-angle resampling and data preprocessing are respectively carried out on the existing electric signals to obtain a plurality of equal-angle dimensionless stable signals irrelevant to the rotating speed, the working condition and the load; respectively demodulating a plurality of the equal-angle dimensionless steady signals and extracting a plurality of fault characteristic indexes by adopting a signal time-frequency analysis method; taking a plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis, and outputting a fault mode of the traction motor and a severity degree corresponding to the fault mode; and carrying out health management on the traction motor according to the plurality of fault characteristic indexes.
Optionally, the existing electrical signals include, but are not limited to, flux linkage signals, inverter current signals, intermediate voltage signals, rotational speed signals, torque signals, motor temperature signals, and pulse signals.
Optionally, the performing equal-angle resampling and data preprocessing on the plurality of existing electrical signals respectively to obtain a plurality of equal-angle dimensionless stationary signals unrelated to the rotation speed, the working condition and the load includes: respectively carrying out equal-angle resampling on the existing electric signals by adopting a step ratio tracking technology to obtain a plurality of corresponding equal-angle stable signals; decomposing and reconstructing the equal-angle stationary signals by adopting a method of combining variational modal decomposition and wiener filtering; and carrying out normalization processing on the reconstructed plurality of equiangular stationary signals to obtain a plurality of equiangular dimensionless stationary signals irrelevant to the rotating speed, the working condition and the load.
Optionally, the decomposing and reconstructing the plurality of equiangular stationary signals by using a method of combining variational modal decomposition and wiener filtering includes: for any one of the equiangular stationary signals, carrying out variation modal decomposition on the equiangular stationary signal to obtain a plurality of modal components corresponding to the equiangular stationary signal; eliminating fundamental frequency and harmonic wave components of each modal component by adopting a wiener filter; and superposing the filtered modal components to obtain the reconstructed equal-angle stationary signal.
Optionally, the demodulating the plurality of equal-angle dimensionless stationary signals and extracting a plurality of fault feature indicators by using a signal time-frequency analysis method respectively includes: respectively demodulating the normalized plurality of equiangular dimensionless stationary signals by adopting a demodulation algorithm to obtain a plurality of corresponding demodulated signals, wherein the demodulation algorithm includes but is not limited to: linear operator demodulation algorithm, synchronous demodulation algorithm and square demodulation algorithm; extracting the characteristic frequency amplitude of at least one fault of the traction motor from the frequency spectrums of the demodulation signals, calculating a plurality of time domain indexes from the time domain sequence of each demodulation signal, and combining the analog indexes and the trend indexes to jointly form a plurality of fault characteristic indexes of the traction motor.
Optionally, the multiple-input multiple-output diagnosis model is an improved gradient lifting tree diagnosis model based on a grey wolf pack, and before the multiple fault characteristic indexes are used as inputs, a preset multiple-input multiple-output diagnosis model is applied to perform fault diagnosis, and a fault mode of the traction motor and a severity corresponding to the fault mode are output, the method includes: acquiring historical fault characteristic indexes serving as model training samples, and fault modes and severity degrees corresponding to the historical fault characteristic indexes serving as true values to form a model training set, and initializing model parameters of the improved gradient lifting tree diagnosis model based on the grey wolf cluster; inputting the historical fault characteristic indexes serving as model training samples into the improved gradient lifting tree diagnosis model based on the gray wolf group for processing, outputting estimated values corresponding to the historical fault characteristic indexes, and calculating the difference between the true values and the estimated values; if the difference is not less than a preset value and the parameter optimization iteration times are not up to the maximum iteration times, carrying out model parameter optimization on the improved gradient lifting tree diagnosis model based on the gray wolf colony, and carrying out repeated training according to the model training set; and if the difference is smaller than a preset value or the parameter optimization iteration number reaches the maximum iteration number, finishing the training of the improved gradient lifting tree diagnosis model based on the grey wolf colony, and obtaining model parameters of the trained improved gradient lifting tree diagnosis model based on the grey wolf colony.
Optionally, the health management of the traction motor according to the plurality of fault characteristic indicators includes: according to the fault characteristic indexes, health state assessment and service life prediction are carried out on the whole and the subcomponents of the traction motor by adopting a hierarchical fuzzy core clustering comprehensive assessment algorithm; and providing a maintenance decision suggestion for the operation and maintenance of the traction motor according to the health state evaluation result.
Based on the same inventive concept, the embodiment of the invention also provides a traction motor health diagnosis system, which comprises: the signal acquisition and preprocessing unit is used for acquiring a plurality of existing electric signals of the traction motor from a traction control unit of the converter, and respectively carrying out equal-angle resampling and data preprocessing on the existing electric signals to obtain a plurality of equal-angle dimensionless steady signals irrelevant to the rotating speed, the working condition and the load; the fault feature extraction unit is used for demodulating the equal-angle dimensionless steady signals respectively and extracting a plurality of fault feature indexes by adopting a signal time-frequency analysis method; the fault diagnosis unit is used for taking the plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis and outputting a fault mode of the traction motor and a severity degree corresponding to the fault mode; and the health management unit is used for carrying out health management on the traction motor according to the plurality of fault characteristic indexes.
Based on the same inventive concept, an embodiment of the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the foregoing method when executing the computer program.
Based on the same inventive concept, the embodiment of the present invention further provides a computer storage medium, in which at least one executable instruction is stored, and the executable instruction causes a processor to execute the foregoing method.
The beneficial effects of the invention are: from the foregoing, embodiments of the present invention provide a method and a system for diagnosing health of a traction motor, where the method includes: the method comprises the steps that a plurality of existing electric signals of a traction motor are obtained from a traction control unit of a converter, and equal-angle resampling and data preprocessing are respectively carried out on the existing electric signals to obtain a plurality of equal-angle dimensionless stable signals irrelevant to the rotating speed, the working condition and the load; respectively demodulating a plurality of the equal-angle dimensionless steady signals and extracting a plurality of fault characteristic indexes by adopting a signal time-frequency analysis method; taking a plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis, and outputting a fault mode of the traction motor and a severity degree corresponding to the fault mode; the traction motor is subjected to health management according to the plurality of fault characteristic indexes, diagnosis errors caused by transient processes such as power supply harmonic waves, variable rotating speed and variable load can be effectively eliminated, and accuracy of non-stable weak signal characteristic extraction and fault diagnosis is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic diagram of a traction motor health diagnostic system in an embodiment of the present invention;
FIG. 2 is a flow chart schematic of a traction motor health diagnostic method in an embodiment of the present invention;
FIG. 3 is a training diagram of an improved gradient-boosted tree diagnostic model based on gray wolf clusters according to an embodiment of the present invention;
fig. 4 is a schematic diagram of an electronic device in an embodiment of the invention.
Detailed Description
To make the objects, technical solutions and advantages of the present disclosure more apparent, the present disclosure will be described in further detail below with reference to specific embodiments and the accompanying drawings.
It is to be noted that technical terms or scientific terms used in the embodiments of the present invention should have a general meaning as understood by one having ordinary skill in the art to which the present disclosure belongs, unless otherwise defined. The use of "first," "second," and similar language in the embodiments of the present invention does not denote any order, quantity, or importance, but rather the terms "first," "second," and similar language are used to distinguish one element from another. The word "comprising" or "comprises", and the like, means that the element or item listed before the word covers the element or item listed after the word and its equivalents, but does not exclude other elements or items. The terms "connected" or "coupled" and the like are not restricted to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "upper", "lower", "left", "right", and the like are used merely to indicate relative positional relationships, and when the absolute position of the object being described is changed, the relative positional relationships may also be changed accordingly.
The embodiment of the invention provides a traction motor health diagnosis system. The traction motor Health diagnosis system provided by the embodiment of the invention is applied to a fault Prediction and Health Management (PHM) system. As shown in fig. 1, the fault prediction and health management system includes a vehicle-mounted traction motor health diagnosis system, a vehicle-ground transmission module, and a ground traction motor health diagnosis system. The traction motor health diagnosis system of the embodiment of the invention can be a vehicle-mounted traction motor health diagnosis system or a ground traction motor health diagnosis system. According to the functional module division, the traction motor health diagnosis system comprises: the system comprises a signal acquisition and preprocessing unit, a fault feature extraction unit, a fault diagnosis unit and a health management unit.
The signal acquisition and preprocessing unit is used for acquiring a plurality of existing electric signals of the traction motor from a Traction Control Unit (TCU) of the converter, and performing equal-angle resampling and data preprocessing on the existing electric signals respectively to obtain a plurality of equal-angle dimensionless stable signals irrelevant to the rotating speed, the working condition and the load. And the fault feature extraction unit is used for demodulating the equal-angle dimensionless steady signals and extracting a plurality of fault feature indexes by adopting a signal time-frequency analysis method. And the fault diagnosis unit is used for taking the plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis and outputting a fault mode of the traction motor and a severity degree corresponding to the fault mode. And the health management unit is used for carrying out health management on the traction motor according to the plurality of fault characteristic indexes.
In the embodiment of the invention, the signal acquisition and preprocessing unit can acquire an equiangular dimensionless steady signal irrelevant to the rotating speed, the working condition and the load, effectively eliminates the influence of actual service conditions such as frequent working condition conversion, strong rotating speed time variation, load disturbance and the like, and specifically comprises a signal acquisition module, a rotating speed tracking module, a self-adaptive noise elimination module and a normalization module. The signal acquisition module can firstly acquire existing electric signals such as flux linkage, inversion current, intermediate voltage, motor rotating speed, motor torque, motor temperature, pulse signals and the like from a Traction Control Unit (TCU) of the converter; the rotating speed tracking module carries out equal-angle resampling on the non-stationary signals by adopting an order ratio tracking technology to obtain equal-angle stationary signals; the adaptive noise elimination module adopts a method of combining variational modal decomposition and wiener filtering to decompose and reconstruct the equiangular stationary signals, so as to realize adaptive noise elimination of strong noise and power supply harmonic signal components; and the normalization module is used for performing normalization processing on the signals by combining the running condition information of the traction motor such as the rotating speed and the torque, removing the influence of load change on the amplitude of the signals and finally obtaining a plurality of equiangular dimensionless stable signals.
The fault feature extraction unit is used for analyzing the equiangular dimensionless steady signals and extracting fault feature indexes and comprises a signal analysis module and a feature index extraction module. The signal analysis module demodulates a plurality of equiangular dimensionless steady signals by adopting demodulation algorithms such as linear operator demodulation, synchronous demodulation, square demodulation and the like, so that effective extraction of fault signals is realized, and a plurality of corresponding demodulated signals are obtained; the characteristic index extraction module adopts a signal time-frequency analysis method to extract a plurality of fault characteristic indexes such as characteristic frequency amplitude, rotor frequency, harmonic amplitude, imbalance index, working condition parameter, analogy index and trend index.
The fault diagnosis unit is used for fault diagnosis and early fault early warning and comprises a model training module and a fault recognition module. The model training module establishes a multi-input multi-output diagnostic model based on a gradient lifting tree by adopting fault characteristic indexes obtained by the signal characteristic extraction unit, and optimizes parameters of the diagnostic model by adopting a grey wolf group optimization algorithm to realize self-learning of the diagnostic model; the fault identification module is used for identifying fault modes and severity of insulation of a traction motor stator winding, rotor broken bars, unbalance, misalignment, air gap unevenness, mass center deviation, bearing pitting or abrasion, gear breakage or gluing and the like. In addition, for single failure mode output, when the output result is in the first threshold range, the mode failure does not exist; when the result is in the second threshold range, indicating that the mode fault exists and the mode fault is in an early stage, and performing fault early warning; and when the result is in a third threshold range, indicating that the mode fault exists and is in a serious stage, and performing fault alarm.
The health management unit is used for health assessment and life prediction of the whole traction motor and the sub-components, and provides maintenance decision suggestions and comprises a health assessment module and an operation and maintenance decision module. Sub-components include, but are not limited to, stator winding insulation, rotors, bearings, and loads. The health evaluation module can realize the health state evaluation and the service life prediction of the whole and the subcomponents (stator winding insulation, rotor, bearing and load) of the traction motor by adopting a hierarchical fuzzy core clustering comprehensive evaluation algorithm according to the current fault characteristic index; and the operation and maintenance decision module provides a maintenance decision suggestion for the intelligent operation and maintenance of the traction motor according to the output result of the health evaluation module.
The embodiment of the invention is characterized in that a traction motor health diagnosis system is respectively arranged on a vehicle and the ground and comprises a signal acquisition and preprocessing unit, a fault feature extraction unit, a fault diagnosis unit and a health management unit, so that the health diagnosis of the traction motor is realized, namely, under the condition of not additionally adding a sensor, the existing electric signals of a traction control unit are fully utilized, the signal processing methods such as equal-angle resampling, variational modal decomposition, wiener filtering and demodulation algorithms are fused, the fault feature indexes of the traction motor irrelevant to the rotating speed and the load are extracted, a gradient lifting tree fault diagnosis model based on grey wolf colony optimization is further called, and the diagnosis and the identification of the severity of common faults such as insulation of a stator winding, broken bars of a rotor, misalignment, unbalance, broken teeth of a gear and the like of the traction motor are realized; the method can effectively eliminate the diagnosis errors caused by transient processes such as PWM power supply harmonic waves, variable rotating speed and variable load, and improve the accuracy of extraction of non-stable weak signal characteristics and fault mode identification.
For convenience of description, the above system is described with functions divided into various modules, which are described separately. Of course, the functions of the modules may be implemented in the same or multiple software and/or hardware in implementing embodiments of the invention.
The system of the above embodiment is applied to a traction motor health diagnosis method described later, and has the beneficial effects of the corresponding method embodiment.
The embodiment of the invention also provides a traction motor health diagnosis method. The traction motor Health diagnosis method provided by the embodiment of the invention is applied to a fault Prediction and Health Management (PHM) system. As shown in fig. 2, the traction motor health diagnosis method includes:
step S11: the method comprises the steps of obtaining a plurality of existing electric signals of a traction motor from a traction control unit of a converter, and conducting equal-angle resampling and data preprocessing on the existing electric signals respectively to obtain a plurality of equal-angle dimensionless stable signals irrelevant to the rotating speed, the working condition and the load.
In an embodiment of the invention, a plurality of existing electrical signals of the traction motor are obtained from a traction control unit of the inverter. The existing electrical signals of the converter include, but are not limited to, flux linkage signals, inverter current signals, intermediate voltage signals, motor speed signals, motor torque signals, motor temperature signals, and pulse signals.
In step S11, first, an order ratio tracking technique is adopted to perform equal-angle resampling on the existing electrical signals, so as to obtain a plurality of corresponding equal-angle stationary signals. For each existing electric signal, a Computed Order Tracking (COT) algorithm is adopted to perform Order resampling, that is, firstly, a rotation speed signal is utilized to solve an angle domain resampling time point of each existing electric signal, then, based on an original value of each existing electric signal, an interpolation method is utilized to obtain an amplitude value at the angle domain resampling time point, and a required equal angle resampling sequence is obtained, namely, an equal angle stable signal corresponding to the existing electric signal.
And then, decomposing and reconstructing the plurality of equal-angle stationary signals respectively by adopting a method of combining variational modal decomposition and wiener filtering. Optionally, for any one of the equiangular stationary signals, performing variational modal decomposition on the equiangular stationary signal to obtain a plurality of modal components corresponding to the equiangular stationary signal; eliminating fundamental frequency and harmonic component of each modal component by adopting a wiener filter; and superposing the filtered modal components to obtain the reconstructed equal-angle stable signal, so that the self-adaptive noise elimination of the strong noise and power supply harmonic signal components is realized, and the weak fault characteristic signal can be effectively enhanced. For example, for the angular stationary signals such as the inverter current signal, the flux linkage signal, the rotating speed signal and the like, a variation modal decomposition algorithm is respectively adopted to obtain modal components of the angular stationary signals such as the inverter current signal, the flux linkage signal, the rotating speed signal and the like, and the modal components of different signals are different. The embodiment of the invention mainly carries out filtering processing on a plurality of modal components corresponding to the inversion current signal, the flux linkage signal and the rotating speed signal, the fundamental frequency and harmonic components of different signals are different in size, and the harmonic component of 1 to 40 times of the fundamental frequency is generally eliminated.
And normalizing the reconstructed plurality of equiangular stationary signals to obtain a plurality of equiangular dimensionless stationary signals irrelevant to the rotating speed, the working condition and the load, and removing the influence of load change on the amplitude of the signals. The embodiment of the invention also performs normalization processing on the equiangular stationary signals such as the torque signal, the inversion current signal, the flux linkage signal and the like, and particularly takes the inversion current as an example and the flux linkage signal is similar. Firstly, calculating an effective value of the inverted current signal, and then dividing an original signal of the inverted current signal by the effective value to obtain an equiangular dimensionless stable signal after the inverted current signal is normalized.
Step S12: and respectively demodulating a plurality of the equal-angle dimensionless steady signals and extracting a plurality of fault characteristic indexes by adopting a signal time-frequency analysis method.
In the embodiment of the present invention, firstly, a plurality of normalized equal-angle dimensionless stationary signals are demodulated by using a demodulation algorithm to obtain a plurality of corresponding demodulated signals, where the demodulation algorithm includes, but is not limited to: linear operator demodulation algorithm, synchronous demodulation algorithm and square demodulation algorithm. In the embodiment of the invention, each fault has corresponding fault characteristic frequency, if the centering on the fault characteristic frequency is 2 times, the amplitude value at the fault characteristic frequency in the equiangular dimensionless stable signal is extracted by adopting a signal demodulation algorithm, and a corresponding demodulated signal is obtained.
Then, each demodulated signal is analyzed by adopting a signal time-frequency analysis method, and a fault characteristic index is obtained. Optionally, the characteristic frequency amplitude of at least one fault of the traction motor is extracted from the frequency spectrums of the plurality of demodulated signals, a plurality of time domain indexes are calculated from the time domain sequence of each demodulated signal, and the plurality of fault characteristic indexes of the traction motor are jointly formed by combining the analog indexes and the trend indexes. Traction motors include, but are not limited to, traction motor misalignment, stator winding insulation, rotor stripping, bearings, gear teeth breakage. Time domain indicators include, but are not limited to, imbalance coefficients, positive sequence components, negative sequence components, and the like. It should be noted that, a plurality of fault characteristic indicators are a plurality of characteristic values at the same time, such as: the amplitude at the fault characteristic frequency of the misalignment, the insulation of the stator winding, the broken rotor bar, the broken bearing and the broken gear of the gear, the unbalance coefficient and the phase of the inverter current and the like.
Step S13: and taking the plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis, and outputting all faults existing in the traction motor and severity degrees corresponding to the faults.
In the embodiment of the invention, the multi-input multi-output diagnostic model is an improved gradient lifting tree diagnostic model based on a grey wolf group. The input of the improved gradient-boosted tree diagnostic model based on the gray wolf colony is a plurality of fault characteristic indexes obtained in step S12, such as characteristic frequency amplitude, rotor frequency and harmonic amplitude, imbalance index, working condition parameter, analogy index and trend index. The output of the improved gradient lifting tree diagnosis model based on the grey wolf group is faults of insulation of a stator winding of the traction motor, broken bars of a rotor, unbalance, misalignment, broken teeth of a gear and the like and corresponding severity. For a single fault and the severity thereof, when the output result is in a first threshold range, the fault is not existed; when the result is in the second threshold range, indicating that the fault exists and is in an early stage, and performing fault early warning; and when the result is in the third threshold range, indicating that the fault exists and is in a serious stage, and performing fault alarm.
Before step S13, model training needs to be performed on the improved gradient modified boosted tree diagnostic model based on the grey wolf pack by using the historical data to optimize model parameters, and model parameters for convergence of the improved gradient boosted tree diagnostic model based on the grey wolf pack are obtained. Optionally, obtaining a historical fault characteristic index as a model training sample, and a fault mode and a severity corresponding to the historical fault characteristic index as a true value, forming a model training set, and initializing a model parameter of the improved gradient lifting tree diagnosis model based on the gray wolf colony; inputting the historical fault characteristic indexes serving as model training samples into the improved gradient lifting tree diagnosis model based on the gray wolf group for processing, outputting estimated values corresponding to the historical fault characteristic indexes, and calculating the difference between the true values and the estimated values; if the difference value is not smaller than a preset value and the parameter optimization iteration times are not up to the maximum iteration times, carrying out model parameter optimization on the improved gradient lifting tree diagnosis model based on the grey wolf cluster, and carrying out repeated training according to the model training set; and if the difference is smaller than a preset value or the parameter optimization iteration number reaches the maximum iteration number, finishing the training of the improved gradient lifting tree diagnosis model based on the grey wolf colony, and obtaining model parameters of the trained improved gradient lifting tree diagnosis model based on the grey wolf colony. As shown in fig. 3 in particular, the inputs of the improved gradient-boosted tree diagnostic model based on the gray wolf population include: characteristic frequency amplitude, rotor frequency and harmonic amplitude, imbalance indicator, operating condition parameter, analogy indicator and trend indicator. The output includes: stator winding insulation, rotor bar breakage, imbalance, misalignment, bearing (outer ring/inner ring/rolling body/cage) and gear tooth breakage, and the corresponding severity. Firstly, initializing model parameters, and then optimizing gradient lifting tree parameters based on a gray wolf group algorithm, wherein the method comprises the following steps: learning step size, iteration number, maximum depth, etc. And then, obtaining an estimated value by using a gradient lifting tree algorithm, and calculating a difference value between the estimated value and a true value. And judging whether the difference value is smaller than a preset value or reaches the maximum iteration number. And if the difference value is not less than the preset value and reaches the maximum iteration times, returning to the step of optimizing the gradient lifting tree parameters based on the gray wolf colony algorithm, otherwise, finishing the model training and obtaining the model parameters of the convergence of the improved gradient lifting tree diagnosis model based on the gray wolf colony.
In step S13, according to the model parameters converged by the improved gradient spanning tree diagnostic model based on the gray wolf group, a plurality of fault feature indicators are input into the improved gradient spanning tree diagnostic model based on the gray wolf group for processing, and all faults existing in the traction motor and the severity corresponding to each fault are output.
Step S14: and carrying out health management on the traction motor according to the plurality of fault characteristic indexes.
In the embodiment of the invention, optionally, a hierarchical fuzzy core clustering comprehensive evaluation algorithm is adopted to evaluate the health state and predict the service life of the whole and the subcomponents of the traction motor according to a plurality of fault characteristic indexes; including but not limited to stator winding insulation, rotors, bearings, loads, etc. And providing maintenance decision suggestions for the operation and maintenance of the traction motor according to the health state evaluation result, such as available remaining time suggestions, maintenance suggestions and the like.
According to the traction motor health diagnosis method, under the condition that no additional sensor is added, the existing electric signals of the traction control unit are fully utilized, the signal processing methods such as equal-angle resampling, variational modal decomposition, wiener filtering and demodulation algorithms are fused, the fault characteristic indexes of the traction motor irrelevant to the rotating speed and the load are extracted, a gradient lifting tree fault diagnosis model based on gray wolf group optimization is further called, fault diagnosis and severity identification of insulation of a stator winding, rotor broken bars, unbalance, misalignment, gear broken teeth and the like of the traction motor are achieved, the effects of extraction and identification of weak and non-stable fault characteristic indexes in the traction fault state can be improved, and fault diagnosis and early warning of insulation of the stator winding, rotor broken bars, unbalance, misalignment, gear broken teeth and the like of the traction motor are achieved.
According to the traction motor health diagnosis method, a plurality of existing electric signals of a traction motor are obtained from a traction control unit of a converter, and equal-angle resampling and data preprocessing are respectively carried out on the existing electric signals, so that a plurality of equal-angle dimensionless stable signals irrelevant to the rotating speed, the working condition and the load are obtained; respectively demodulating the plurality of equiangular dimensionless steady signals and extracting a plurality of fault characteristic indexes by adopting a signal time-frequency analysis method; taking a plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis, and outputting all faults existing in the traction motor and severity corresponding to each fault; the traction motor is subjected to health management according to the plurality of fault characteristic indexes, diagnosis errors caused by transient processes such as power supply harmonic waves, variable rotating speed and variable load can be effectively eliminated, and accuracy of non-stable weak signal characteristic extraction and fault diagnosis is improved.
The foregoing description of specific embodiments of the present invention has been presented. In some cases, acts or steps recited in embodiments of the invention may be performed in an order different than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
Based on the same inventive concept, embodiments of the present invention further provide an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the method according to any of the above embodiments is implemented.
An embodiment of the present invention provides a non-volatile computer storage medium, where at least one executable instruction is stored in the computer storage medium, and the computer executable instruction may execute the method described in any of the above embodiments.
Fig. 4 is a schematic diagram illustrating a more specific hardware structure of an electronic device according to this embodiment, where the device may include: a processor 401, a memory 402, an input/output interface 403, a communication interface 404, and a bus 405. Wherein the processor 401, the memory 402, the input/output interface 403 and the communication interface 404 are communicatively connected to each other within the device by a bus 405.
The processor 401 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits, and is configured to execute related programs to implement the technical solution provided by the embodiment of the present invention.
The Memory 402 may be implemented in the form of a ROM (Read Only Memory), a RAM (Random access Memory), a static storage device, a dynamic storage device, or the like. The memory 402 may store an operating system and other application programs, and when the technical solution provided by the method embodiment of the present invention is implemented by software or firmware, the relevant program codes are stored in the memory 402 and called to be executed by the processor 401.
The input/output interface 403 is used for connecting an input/output module to realize information input and output. The i/o module may be configured as a component in a device (not shown) or may be external to the device to provide a corresponding function. Wherein the input devices may include a keyboard, mouse, touch screen, microphone, various sensors, etc., and the output devices may include a display, speaker, vibrator, indicator light, etc.
The communication interface 404 is used to connect a communication module (not shown in the figure) to implement communication interaction between the present device and other devices. The communication module can realize communication in a wired mode (such as USB, network cable and the like) and also can realize communication in a wireless mode (such as mobile network, WIFI, bluetooth and the like).
The bus 405 includes a path that transfers information between the various components of the device, such as the processor 401, memory 402, input/output interface 403, and communication interface 404.
It should be noted that although the above-mentioned device only shows the processor 401, the memory 402, the input/output interface 403, the communication interface 404 and the bus 405, in a specific implementation, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will appreciate that the above-described apparatus may also include only those components necessary to implement embodiments of the present invention, and need not include all of the components shown in the figures.
Those of ordinary skill in the art will understand that: the discussion of any embodiment above is meant to be exemplary only, and is not intended to intimate that the scope of the disclosure is limited to these examples; within the context of the present application, features from the above embodiments or from different embodiments may also be combined, steps may be implemented in any order, and there are many other variations of the different aspects of the present application as described above, which are not provided in detail for the sake of brevity.
This application is intended to cover all such alternatives, modifications and variations that fall within the broad scope of embodiments of the present invention. Therefore, any omissions, modifications, substitutions, improvements, and the like that may be made without departing from the spirit and principles of the embodiments of the present invention are intended to be included within the scope of the present application.
Claims (10)
1. A traction motor health diagnostic method, the method comprising:
the method comprises the steps that a plurality of existing electric signals of a traction motor are obtained from a traction control unit of a converter, and equal-angle resampling and data preprocessing are respectively carried out on the existing electric signals to obtain a plurality of equal-angle dimensionless stable signals irrelevant to the rotating speed, the working condition and the load;
respectively demodulating a plurality of the equal-angle dimensionless steady signals and extracting a plurality of fault characteristic indexes by adopting a signal time-frequency analysis method;
taking a plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis, and outputting all faults existing in the traction motor and severity corresponding to each fault;
and carrying out health management on the traction motor according to the plurality of fault characteristic indexes.
2. The method of claim 1, wherein the existing electrical signals include, but are not limited to, flux linkage signals, inverter current signals, intermediate voltage signals, rotational speed signals, torque signals, motor temperature signals, and pulse signals.
3. The method of claim 1, wherein said performing equal angle resampling and data preprocessing on said plurality of existing electrical signals to obtain a plurality of equal angle dimensionless stationary signals independent of speed, condition and load comprises:
carrying out equal-angle resampling on the existing electric signals by adopting an order ratio tracking technology to obtain a plurality of corresponding equal-angle stable signals;
respectively decomposing and reconstructing the equiangular stationary signals by adopting a method of combining variational modal decomposition and wiener filtering;
and carrying out normalization processing on the reconstructed plurality of equiangular stationary signals to obtain a plurality of equiangular dimensionless stationary signals irrelevant to the rotating speed, the working condition and the load.
4. The method as claimed in claim 3, wherein the decomposing and reconstructing the plurality of equiangular stationary signals by using the method of combining the variational modal decomposition and the wiener filtering comprises: for any of the said equal angle stationary signals,
carrying out variation modal decomposition on the equal-angle stationary signal to obtain a plurality of modal components corresponding to the equal-angle stationary signal;
eliminating fundamental frequency and harmonic component of each modal component by adopting a wiener filter;
and superposing the filtered modal components to obtain the reconstructed equal-angle stationary signal.
5. The method as claimed in claim 1, wherein said demodulating a plurality of said equiangular dimensionless stationary signals and extracting a plurality of fault feature indicators by signal time-frequency analysis, respectively, comprises:
respectively demodulating the normalized plurality of equal-angle dimensionless stationary signals by adopting a demodulation algorithm to obtain a plurality of corresponding demodulated signals, wherein the demodulation algorithm includes but is not limited to: linear operator demodulation algorithm, synchronous demodulation algorithm and square demodulation algorithm;
extracting the characteristic frequency amplitude of at least one fault of the traction motor from the frequency spectrums of the plurality of demodulation signals, calculating a plurality of time domain indexes from the time domain sequence of each demodulation signal, and combining the analog indexes and the trend indexes to jointly form a plurality of fault characteristic indexes of the traction motor.
6. The method of claim 1, wherein the multiple-input multiple-output diagnostic model is a gray wolf pack-based modified gradient spanning tree diagnostic model, and wherein the step of performing fault diagnosis by using a preset multiple-input multiple-output diagnostic model with a plurality of fault characteristic indicators as inputs and outputting a fault pattern of the traction motor and a severity corresponding to the fault pattern comprises:
acquiring historical fault characteristic indexes serving as model training samples, and fault modes and severity which correspond to the historical fault characteristic indexes and serve as true values, forming a model training set, and initializing model parameters of the improved gradient lifting tree diagnosis model based on the gray wolf colony;
inputting the historical fault characteristic indexes serving as model training samples into the improved gradient lifting tree diagnosis model based on the grey wolf group for processing, outputting an estimated value corresponding to the historical fault characteristic indexes, and calculating a difference value between the true value and the estimated value;
if the difference is not less than a preset value and the parameter optimization iteration times are not up to the maximum iteration times, carrying out model parameter optimization on the improved gradient lifting tree diagnosis model based on the gray wolf colony, and carrying out repeated training according to the model training set; and if the difference is smaller than a preset value or the parameter optimization iteration number reaches the maximum iteration number, finishing the training of the improved gradient lifting tree diagnosis model based on the grey wolf colony, and obtaining model parameters of the trained improved gradient lifting tree diagnosis model based on the grey wolf colony.
7. The method of claim 1, wherein said health managing traction motors based on a plurality of said fault signature indicators comprises:
according to the fault characteristic indexes, health state assessment and service life prediction are carried out on the whole and the subcomponents of the traction motor by adopting a hierarchical fuzzy core clustering comprehensive assessment algorithm;
and providing a maintenance decision suggestion for the operation and maintenance of the traction motor according to the health state evaluation result.
8. A traction motor health diagnostic system, the system comprising:
the signal acquisition and preprocessing unit is used for acquiring a plurality of existing electric signals of the traction motor from a traction control unit of the converter, and performing equal-angle resampling and data preprocessing on the existing electric signals respectively to obtain a plurality of equal-angle dimensionless stable signals irrelevant to the rotating speed, the working condition and the load;
the fault feature extraction unit is used for demodulating the equal-angle dimensionless steady signals respectively and extracting a plurality of fault feature indexes by adopting a signal time-frequency analysis method;
the fault diagnosis unit is used for taking the plurality of fault characteristic indexes as input, applying a preset multi-input multi-output diagnosis model to carry out fault diagnosis and outputting a fault mode of the traction motor and a severity degree corresponding to the fault mode;
and the health management unit is used for carrying out health management on the traction motor according to the plurality of fault characteristic indexes.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method according to any of claims 1-7 when executing the program.
10. A computer storage medium having stored therein at least one executable instruction for causing a processor to perform the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210943613.3A CN115329810A (en) | 2022-08-08 | 2022-08-08 | Traction motor health diagnosis method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210943613.3A CN115329810A (en) | 2022-08-08 | 2022-08-08 | Traction motor health diagnosis method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115329810A true CN115329810A (en) | 2022-11-11 |
Family
ID=83922694
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210943613.3A Pending CN115329810A (en) | 2022-08-08 | 2022-08-08 | Traction motor health diagnosis method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115329810A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116184200A (en) * | 2023-04-26 | 2023-05-30 | 国家石油天然气管网集团有限公司 | Health state assessment method and system for induction motor of oil transfer pump |
-
2022
- 2022-08-08 CN CN202210943613.3A patent/CN115329810A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116184200A (en) * | 2023-04-26 | 2023-05-30 | 国家石油天然气管网集团有限公司 | Health state assessment method and system for induction motor of oil transfer pump |
| CN116184200B (en) * | 2023-04-26 | 2023-08-04 | 国家石油天然气管网集团有限公司 | Health state assessment method and system for induction motor of oil transfer pump |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ertunc et al. | ANN-and ANFIS-based multi-staged decision algorithm for the detection and diagnosis of bearing faults | |
| Chen et al. | Generator bearing fault diagnosis for wind turbine via empirical wavelet transform using measured vibration signals | |
| EP3816604A1 (en) | Motor bearing failure diagnosis device | |
| Saidi et al. | Bi-spectrum based-EMD applied to the non-stationary vibration signals for bearing faults diagnosis | |
| CN111208427B (en) | Traction motor fault diagnosis method and device | |
| CN110940917B (en) | Motor fault early warning method and system | |
| CN111365251B (en) | Intelligent diagnosis method for faults of centrifugal pump unit | |
| CN113009334B (en) | Motor fault detection method and system based on wavelet packet energy analysis | |
| CN111881736A (en) | Rolling bearing early fault diagnosis method based on bandwidth Fourier decomposition | |
| CN115329810A (en) | Traction motor health diagnosis method and system | |
| CN112115802A (en) | Crane slewing mechanism gear fault diagnosis method, system and storage medium | |
| CN113654798A (en) | Fault diagnosis method and device and electronic equipment | |
| Soualhi et al. | PHM survey: implementation of signal processing methods for monitoring bearings and gearboxes | |
| Niu et al. | Motor fault diagnostics based on current signatures: a review | |
| Guezam et al. | Fault diagnosis of induction motors rotor using current signature with different signal processing techniques | |
| Wang et al. | Research on gearbox composite fault diagnosis based on improved local mean decomposition | |
| CN111610038B (en) | Fault diagnosis system, method, device, equipment and storage medium thereof | |
| CN115096581B (en) | Complex transmission device fault diagnosis tracing method and system based on time-frequency domain characteristics | |
| CN114742111B (en) | Fault diagnosis method and system based on parameter adaptive characteristic modal decomposition | |
| CN116859240A (en) | Direct-current brushless motor fault identification method based on multiple sensors and edge calculation | |
| Yao et al. | A low-frequency fault detection method for low-speed planetary gearbox based on acoustic signals | |
| CN113138337B (en) | Three-phase motor fault diagnosis method based on state division and frequency band synchronous correction | |
| CN113447267B (en) | Vibration signal analysis-based complete gearbox state evaluation method and system | |
| CN112287775A (en) | Motor fault diagnosis method and device and related components | |
| Ma et al. | Compound fault diagnosis of rolling bearing based on ALIF-KELM |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |