Motor fault diagnosis model training method and device, equipment and storage medium

By building a current model in the motor, performing fault simulation and training a fault diagnosis model, the problem of insufficient data in motor fault diagnosis is solved, the accuracy of fault detection is improved, and the motor can be detected in a timely manner before damage is damaged and fault prevention is prevented.

CN118964997APending Publication Date: 2024-11-15HARBIN SIZHERUI INTELLIGENT MEDICAL EQUIP CO LTD +1

Patent Information

Application Number
CN202411084796.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-08
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

In the prior art, due to insufficient data volume of the motor, the diagnostic accuracy of the motor is low, making it difficult to perform effective fault detection before the motor is damaged.

Method used

By building the current model of the motor, performing fault simulation operations to generate fault sample data, building sample space, and training the initial fault diagnosis model based on this to improve the accuracy of the fault diagnosis model.

Benefits of technology

By simulation generating sufficient fault data, the accuracy of the motor fault diagnosis model is improved, ensuring that the fault can be detected in time before the motor is damaged, and avoid affecting the normal operation of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118964997A_ABST
    Figure CN118964997A_ABST
Patent Text Reader

Abstract

The invention discloses a training method, device and equipment for a fault diagnosis model of a motor and a storage medium, and the method comprises the steps: constructing a current model of the motor, and the current model comprises a plurality of physical parameters in the motor; in response to a fault simulation operation on the current model, changing a parameter value of at least one physical parameter of the current model, and operating the changed current model to obtain fault sample data corresponding to the fault simulation operation; constructing a sample space of the motor based on the fault sample data; and constructing an initial fault diagnosis model, and training the initial fault diagnosis model based on the sample space to obtain a fault diagnosis model of the motor. According to the invention, fault simulation is carried out on the current model of the motor through the fault simulation operation to obtain the fault sample data, so that the problem of inaccurate fault diagnosis model caused by insufficient data volume is solved, and the accuracy of fault diagnosis of the fault diagnosis model of the motor is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to a training method, device, equipment and storage medium for a fault diagnosis model of a motor. Background Art

[0002] As a common power device, the motor has the advantages of long service life and good stability, and has been widely used in aerospace, medical equipment and other fields. However, due to factors such as sudden load changes and harsh working environment, the motor often fails and is damaged. Once the motor is damaged, it will seriously affect the normal operation of the system. Taking the medical surgical robot as an example, the joints of the medical surgical robot are usually driven by a motor. When the motor works normally, the joints of the robot can complete rotation or movement under the drive of the motor. The combination of the movements of each joint enables the medical surgical robot to complete complex surgical movements. When the motor is damaged due to a fault, the movement of the corresponding joint will be uncontrollable, which will affect the normal operation of the surgical robot, affect the surgical process, and even endanger the patient's life in severe cases. For this reason, how to diagnose the fault of the motor before the motor is damaged has always been a hot issue in the field of fault diagnosis.

[0003] At present, a method combining measured data with machine learning is usually used for motor fault diagnosis. However, in actual situations, due to the small number of data samples when a motor fails, there is a large difference in the amount of data between the normal state and the abnormal state of the motor. Therefore, how to obtain sufficient motor abnormality data has become a bottleneck problem that limits the accuracy of motor fault diagnosis. To this end, the present invention proposes a numerical model-driven motor fault diagnosis method. The method generates sufficient motor abnormality data by establishing a numerical model of the motor current, solving the problem of inaccurate fault diagnosis model due to insufficient data. Summary of the invention

[0004] The present invention provides a method, device, equipment and storage medium for training a fault diagnosis model of a motor to solve the problem of inaccurate fault diagnosis model due to insufficient data.

[0005] According to one aspect of the present invention, a method for training a fault diagnosis model of a motor is provided, comprising:

[0006] Constructing a current model of the motor, wherein the current model includes multiple physical parameters inside the motor;

[0007] In response to a fault simulation operation on the current model, a parameter value of at least one physical parameter of the current model is changed, and the changed current model is run to obtain fault sample data corresponding to the fault simulation operation;

[0008] Constructing a sample space of the motor based on the fault sample data;

[0009] An initial fault diagnosis model is constructed, and the initial fault diagnosis model is trained based on the sample space to obtain a fault diagnosis model of the motor.

[0010] According to another aspect of the present invention, a method for diagnosing a fault of a motor is provided, comprising:

[0011] Acquire the current signal of the motor to be diagnosed at the current moment, and extract characteristic data of the current signal;

[0012] The characteristic data of the current signal is input into the fault diagnosis model corresponding to the motor to be diagnosed to perform diagnosis, so as to obtain the fault diagnosis result of the motor to be diagnosed; wherein, the fault diagnosis model is trained based on the training method of the motor fault diagnosis model described in any embodiment of the present invention.

[0013] According to another aspect of the present invention, there is provided a training device for a fault diagnosis model of a motor, comprising:

[0014] A current model building module, used to build a current model of the motor, wherein the current model includes multiple physical parameters inside the motor;

[0015] A fault simulation module, configured to change a parameter value of at least one physical parameter of the current model in response to a fault simulation operation on the current model, and to run the changed current model to obtain fault sample data corresponding to the fault simulation operation;

[0016] A sample space construction module, used to construct a sample space of the motor based on the fault sample data;

[0017] The fault diagnosis model training module is used to construct an initial fault diagnosis model, and train the initial fault diagnosis model based on the sample space to obtain the fault diagnosis model of the motor.

[0018] According to another aspect of the present invention, there is provided a fault diagnosis device for a motor, comprising:

[0019] A current signal acquisition module, used to acquire the current signal of the motor to be diagnosed at the current moment, and extract characteristic data of the current signal;

[0020] A fault diagnosis module is used to input the characteristic data of the current signal into the fault diagnosis model corresponding to the motor to be diagnosed for diagnosis, so as to obtain the fault diagnosis result of the motor to be diagnosed; wherein, the fault diagnosis model is trained based on the training method of the motor fault diagnosis model described in any embodiment of the present invention.

[0021] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0022] at least one processor; and

[0023] a memory communicatively connected to the at least one processor; wherein,

[0024] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the training method of the motor fault diagnosis model described in any embodiment of the present invention, and / or the motor fault diagnosis method described in any embodiment of the present invention.

[0025] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the training method of the motor fault diagnosis model described in any embodiment of the present invention, and / or the motor fault diagnosis method described in any embodiment of the present invention when executed.

[0026] The technical solution of the embodiment of the present invention simulates fault sample data in the motor model through a fault simulation operation, constructs a sample space of the motor based on the fault sample data, and trains an initial fault diagnosis model based on the sample space to obtain a fault diagnosis model of the motor, thereby solving the problem of inaccurate fault diagnosis model caused by insufficient data, thereby improving the accuracy of fault diagnosis of the fault diagnosis model of the motor.

[0027] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 This is a process of a method for training a motor fault diagnosis model provided by the first embodiment of the present invention;

[0030] Figure 2 is a structural schematic diagram of a DC motor provided in Embodiment 1 of the present invention;

[0031] Figure 3is a flow chart of a motor fault diagnosis method provided in Embodiment 2 of the present invention;

[0032] Figure 4 It is a structural schematic diagram of a training device for a motor fault diagnosis model provided by a third embodiment of the present invention;

[0033] Figure 5 It is a structural schematic diagram of a motor fault diagnosis device provided by a fourth embodiment of the present invention;

[0034] Figure 6 It is a structural schematic diagram of an electronic device provided in Embodiment 5 of the present invention. DETAILED DESCRIPTION

[0035] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0036] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0037] Embodiment 1

[0038] Figure 1 This is a flow chart of a method for training a motor fault diagnosis model provided by the first embodiment of the present invention. This embodiment is applicable to the case of training a motor fault detection model. The method can be executed by a training device for a motor fault diagnosis model. The training device for the motor fault diagnosis model can be implemented in the form of hardware and / or software. The training device for the motor fault diagnosis model can be configured in electronic devices such as computers and servers. Figure 1 As shown, the method includes:

[0039] S110 , constructing a current model of the motor, wherein the current model includes a plurality of physical parameters inside the motor.

[0040] The physical parameters include but are not limited to the inductance, resistance, and rotational inertia of the motor rotor, etc., which are not limited here. For example, taking a DC motor as an example, Figure 2 is a schematic diagram of the structure of a DC motor provided in Embodiment 1 of the present invention, such as Figure 2 As shown, the DC motor is composed of an electrical network system and a mechanical motion system, and the current model of the motor includes an electrical network balance sub-model and a mechanical balance sub-model of the motor. Specifically, the electrical network balance sub-model is as follows:

[0041]

[0042] Among them, I a is the armature current of the motor; R a is the motor resistance; L a is the inductance of the motor; E a is the induced electromotive force of the armature winding, which can be expressed as:

[0043] E a =K e ω

[0044] Among them, k e is the electromotive force constant, which is determined by the motor structure parameters; ω is the motor rotation angular velocity.

[0045] The mechanical balance submodel is as follows:

[0046]

[0047] Among them, J a is the moment of inertia of the motor rotor; M a is the electromagnetic torque of the motor; M L is the resistance torque, M when no load L =0. At the same time, the torque balance equation can be expressed as:

[0048] M a =K c I a

[0049] Among them, K c is the electromagnetic torque constant, which is determined by the structural parameters of the motor.

[0050] Combining the above electrical network balance sub-model and mechanical balance sub-model, the current model of the motor is obtained. The current model of the motor is as follows:

[0051]

[0052] Understandably, La , R a , K c , K e and J a It is the internal physical parameter of the motor.

[0053] In some embodiments, optionally, after constructing the current model of the motor, the method further includes: constructing particles in a particle swarm algorithm based on the physical parameters of the current model; randomly initializing a particle position and a particle velocity of each of the particles; iteratively updating the particle position and the particle velocity of each of the particles based on the particle swarm algorithm, and determining an updated fitness value based on a fitness function; when an iteration stop condition is met, taking the particle obtained in the current iteration as the optimal particle, and updating the current model based on the physical parameters constituting the optimal particle.

[0054] In this embodiment, for the constructed current model, a particle swarm algorithm can be used to optimize the physical parameters in the current model to make the current model closer to the real motor; specifically, the optimization process of the particle swarm algorithm is as follows:

[0055] 1) The physical parameters of the current model constitute the particles in the particle swarm algorithm, where each particle represents a set of physical parameters;

[0056] 2) Randomly initialize the particle position and particle velocity of each particle; specifically, for each particle, randomly generate a set of initial values ​​of physical parameters, which constitute the initial position of the particle in the search space. Similarly, randomly assign an initial velocity to each particle, which will guide the moving direction and step size of the particle in the search space.

[0057] 3) Iteratively updating the particle position and particle velocity of each particle based on the particle swarm algorithm, and determining the updated fitness value based on the fitness function;

[0058] Specifically, in each iteration, the particle position and particle speed of each particle are updated according to the particle's own historical optimal position, the historical optimal position of the entire population, and the current speed; for the updated particle position, its fitness value is calculated using a predefined fitness function; wherein the fitness function is used to evaluate each particle.

[0059] 4) When the iteration stopping condition is met, the particle obtained in the current iteration is taken as the optimal particle, and the current model is updated based on the physical parameters constituting the optimal particle.

[0060] Among them, the iteration stop conditions include but are not limited to the number of iterations reaching the maximum number of iterations, the fitness value reaching the preset threshold, the change of the global optimal solution in multiple consecutive iterations is less than the preset threshold, the running time exceeds the preset time threshold, etc., which are not limited here. During the iteration process, when the iteration stop conditions are met, the particle with the highest fitness value in the current iteration is used as the optimal particle. Based on the physical parameters of the optimal particle, the current model is updated. This means that the physical parameter values ​​of the optimal particle are applied to the current model in the hope of obtaining better performance or meeting specific design requirements.

[0061] Based on the above embodiment, optionally, determining the updated fitness value based on the fitness function includes: obtaining the real current signal of the motor in a normal state, and performing noise reduction processing on the real current signal; determining the updated fitness value based on the output current signal of the current model and the real current signal after noise reduction processing.

[0062] Among them, the real current signal refers to the working current of the motor under normal conditions. In this embodiment, in the iterative process, the real current signal of the motor under normal conditions is obtained, and the real current signal can be denoised based on a preset noise reduction algorithm, and the output current signal of the current model and the real current signal after noise reduction are determined based on a preset fitness function to determine the updated fitness value. Among them, the preset noise reduction algorithm digital filter (such as low-pass filter, band-pass filter), wavelet transform, Kalman filter, singular value decomposition (Singular Value Decomposition, SVD), etc. are preset. Specifically, taking the singular value decomposition algorithm to perform noise reduction on the real current signal as an example, assuming that the noisy real current signal is Y, it can be expressed as the superposition of the noise signal D and the noise-free current signal X. The fundamental purpose of using SVD for noise reduction is to find the optimal set of Make Close to the real current signal X, the formula can be expressed as in is the Frobenious norm. It can be decomposed by Y, and its formula can be expressed as:

[0063]

[0064] Among them, u k 、v k are the left and right singular vectors after singular value decomposition of the noisy current signal Y, σ k are the first r largest singular values ​​of the noisy current signal Y, where r<rank(X).

[0065] Exemplarily, the preset fitness function is as follows:

[0066]

[0067] Among them, F is the fitness value, N is the total number of samples; i k is the output current signal of the kth current model in the sample, is the kth real current signal.

[0068] S120. In response to a fault simulation operation on the current model, a parameter value of at least one physical parameter of the current model is changed, and the changed current model is run to obtain fault sample data corresponding to the fault simulation operation.

[0069] Among them, the fault simulation operation is used to simulate different fault conditions. For example, when the motor inductance is damaged, its inductance value will decrease. At this time, the inductance value in the current model can be changed to generate fault sample data corresponding to the damage condition. Fault sample data refers to the current data under the fault condition corresponding to the fault simulation operation.

[0070] Fault sample data refers to the current signal under the fault state corresponding to the fault simulation operation.

[0071] It can be understood that a large amount of fault data under different fault states can be obtained by repeatedly executing step S120.

[0072] S130: construct a sample space of the motor based on the fault sample data.

[0073] The sample space includes feature data of fault sample data under different fault states. In this embodiment, feature extraction and dimensionality reduction processing are performed on the fault sample data under different fault states to construct the sample space of the motor.

[0074] On the basis of the above embodiment, optionally, constructing the sample space of the motor based on the fault sample data includes: performing feature extraction and dimensionality reduction processing on the time domain signal of the fault sample data to obtain feature data corresponding to the fault sample data; and constructing the sample space based on the feature data corresponding to the fault sample data.

[0075] Specifically, feature extraction is performed on the time domain features of the current signal in the fault sample data to obtain feature data corresponding to the fault sample data. Feature data includes but is not limited to the maximum value, minimum value, average value, variance, peak value, etc. of the current signal, which are not limited here.

[0076] It is understandable that sufficient feature data can ensure the robustness of the fault diagnosis model, but too much feature data will affect the accuracy of model training. Therefore, after extracting the feature data, the extracted feature data is subjected to dimensionality reduction processing. Exemplarily, principal component analysis (PCA) can be used to reduce the dimensionality of the feature data. The matrix X containing the n-dimensional features of m fault samples obtained by time domain feature extraction can be expressed as:

[0077]

[0078] The matrix X is centered using the mean value of each feature. The mean value of the feature can be expressed as

[0079]

[0080] Among them, μ i is the average value of the i-th dimension feature on m samples. The matrix X is centralized to obtain a new matrix X0, which is expressed as:

[0081]

[0082] Further, we get the covariance matrix X of X0 COV , the expression is:

[0083]

[0084] Furthermore, the covariance matrix X is obtained based on singular value decomposition COV The characteristic data and corresponding characteristic vector are as follows:

[0085] X cov =UΣV T

[0086] Among them, U is an m*m square matrix, called a left singular matrix; Σ is an m*n real diagonal matrix, and V is an n*n square matrix, called a right singular matrix.

[0087] Arrange the feature data from large to small, select the largest kp, and use the corresponding feature vectors as row vectors to form the feature vector X p , based on the feature matrix X, determine the reduced feature matrix X n , whose formula is: X n =X p X. Construct the sample space based on the feature matrix after dimensionality reduction.

[0088] S140, constructing an initial fault diagnosis model, and training the initial fault diagnosis model based on the sample space to obtain a fault diagnosis model of the motor.

[0089] The initial fault diagnosis model can be a neural network model, such as a BP neural network model, a convolutional neural network, a generative adversarial network, etc., which is not limited here. In this embodiment, an initial fault diagnosis model is constructed, and the initial fault diagnosis model is trained based on the feature data in the sample space to obtain a fault diagnosis model of the motor.

[0090] Based on the above embodiment, optionally, constructing the initial fault diagnosis model includes: establishing a neural network model, and initializing the weights and thresholds of the neural network model; updating the weights and thresholds of the neural network model based on a preset optimization algorithm to obtain the initial fault diagnosis model.

[0091] In this embodiment, a neural network model is established, and the weights and thresholds of the neural network model are initialized; it is understandable that in the neural network, the weights and thresholds (i.e., bias terms) are randomly initialized. Common initialization methods include, but are not limited to, small random number initialization, Xavier / Glorot initialization, He initialization, etc., which are not limited here. Further, the weights and thresholds of the neural network model are updated based on a preset optimization algorithm to obtain an initial fault diagnosis model. Among them, the preset optimization algorithm includes, but is not limited to, gradient descent, batch gradient descent, Grey Wolf Optimizer (GWO), etc., which are not limited here.

[0092] Specifically, taking the optimization of the BP neural network model by the gray wolf optimization algorithm as an example, the process of optimizing the BP neural network model by the gray wolf optimization algorithm is as follows:

[0093] 1) Determine the BP neural network topology;

[0094] It should be noted that the structure of the BP neural network, as the main part of fault detection, has a great influence on the detection results. A large number of research results show that for mechanical fault detection problems, the basic three-layer network structure can meet the detection accuracy requirements. Generally speaking, the calculation formula for the number of hidden layer neuron nodes can be expressed as:

[0095]

[0096] Among them, n in is the number of input layer nodes, and its value is equal to the number of eigenvalues ​​extracted from the sample space; n out is the number of nodes in the output layer; m is a constant, usually 1-10.

[0097] 2) Initialize GWO;

[0098] Specifically, according to the basic structure of the BP neural network, the parameters to be optimized are determined, so as to set the number of gray wolves in GWO and other parameters. The root mean square error (MSE) of the BP neural network prediction results is selected as the fitness function F of GWO, and its expression is:

[0099]

[0100] Among them, F is the fitness; N is the number of neural network training samples; y i and is the expected value and actual value of the neural network. According to the fitness calculation results, select α wolf, β wolf, and δ wolf from the gray wolf group.

[0101] 3) Update of GWO;

[0102] Specifically, the entire GWO update process can be divided into four stages from the perspective of collective hunting:

[0103] (1) Surrounding prey. During hunting, the behavior of gray wolves surrounding prey can be defined as:

[0104]

[0105] in, is the distance vector between the individual and the prey, is the prey position vector, is the gray wolf position vector, and is the coefficient vector, which can be expressed as: is the convergence factor, and its formula is: A random number between 0 and 1.

[0106] (2) Wolf pack movement. This process forces the gray wolf individuals to move around with α wolf, β wolf, and δ wolf as the core. The movement of the gray wolf can be expressed as:

[0107]

[0108] in, and is the distance between the gray wolf individual and α wolf, β wolf and δ wolf; They are represented as the current positions of wolf α, wolf β and wolf δ respectively; is the direction and distance that the gray wolf moves toward wolf α, wolf β, and wolf δ, is the final movement direction and distance of the gray wolf individual.

[0109] (3) Searching for prey. When it drops from 2 to 1, the gray wolf population continues to search for prey. When it is equal to 1, the prey search is completed.

[0110] (4) Attack prey. When it is less than 1, the wolves gradually approach the prey. When it is equal to 0, the wolves have finished attacking their prey.

[0111] 4) Establish initial fault diagnosis model

[0112] The node values ​​of the BP neural network model are updated according to the GWO optimization algorithm in step 3), and the target weights and target thresholds of the BP neural network are obtained. An initial fault diagnosis model is established based on the target weights and target thresholds.

[0113] The technical solution of this embodiment simulates fault sample data in the motor model through a fault simulation operation, constructs a sample space of the motor based on the fault sample data, trains an initial fault diagnosis model based on the sample space, and obtains a fault diagnosis model of the motor, thereby solving the problem of inaccurate fault diagnosis model due to insufficient data and improving the accuracy of fault diagnosis of the fault diagnosis model of the motor.

[0114] Embodiment 2

[0115] Figure 3 This is a flow chart of a motor fault diagnosis method provided by Embodiment 2 of the present invention. Based on the above embodiments, this embodiment performs fault diagnosis on the motor to be diagnosed based on the fault diagnosis model trained by the above embodiments. The explanations of the terms that are the same or corresponding to the above embodiments are not repeated here.

[0116] like Figure 3 As shown, the method includes:

[0117] S210, obtaining a current signal of the motor to be diagnosed at a current moment, and extracting characteristic data of the current signal.

[0118] S220. Input the characteristic data of the current signal into the fault diagnosis model corresponding to the motor to be diagnosed to perform diagnosis, and obtain the fault diagnosis result of the motor to be diagnosed; wherein, the fault diagnosis model is trained based on the training method of the motor fault diagnosis model described in any embodiment of the present invention.

[0119] The motor to be diagnosed refers to the motor to be diagnosed for fault. In this embodiment, the current signal of the motor to be diagnosed at the current moment is obtained, and the characteristic data of the current signal is extracted; the characteristic data of the current signal is input into the fault diagnosis model corresponding to the motor to be diagnosed for diagnosis, and the fault diagnosis result of the motor to be diagnosed is obtained. The fault diagnosis result is a binary classification result, and the fault diagnosis result includes whether the motor to be diagnosed has a fault or not.

[0120] It should be noted that each motor to be diagnosed corresponds to a fault diagnosis model one by one. When performing fault diagnosis on the motor to be diagnosed, the fault diagnosis model corresponding to the motor to be diagnosed needs to be loaded.

[0121] The technical solution of this embodiment obtains the current signal of the motor to be diagnosed at the current moment, extracts the characteristic data of the current signal, and inputs the characteristic data of the current signal into the fault diagnosis model corresponding to the motor to be diagnosed for diagnosis, thereby obtaining the fault diagnosis result of the motor to be diagnosed, which can improve the accuracy of fault diagnosis.

[0122] Embodiment 3

[0123] Figure 4 FIG. 1 is a schematic diagram of a structure of a training device for a motor fault diagnosis model provided by Embodiment 3 of the present invention. Figure 4 As shown, the device comprises:

[0124] A current model building module 310 is used to build a current model of the motor, wherein the current model includes multiple physical parameters inside the motor;

[0125] A fault simulation module 320, configured to change a parameter value of at least one physical parameter of the current model in response to a fault simulation operation on the current model, and to run the changed current model to obtain fault sample data corresponding to the fault simulation operation;

[0126] A sample space construction module 330, configured to construct a sample space of the motor based on the fault sample data;

[0127] The fault diagnosis model training module 340 is used to construct an initial fault diagnosis model, and train the initial fault diagnosis model based on the sample space to obtain the fault diagnosis model of the motor.

[0128] The technical solution of this embodiment simulates fault sample data in the motor model through a fault simulation operation, constructs a sample space of the motor based on the fault sample data, and trains an initial fault diagnosis model based on the sample space to obtain a fault diagnosis model of the motor, thereby solving the problem of inaccurate fault diagnosis model due to insufficient data, thereby improving the accuracy of fault diagnosis of the fault diagnosis model of the motor.

[0129] Based on the above embodiment, optionally, the current model of the motor includes an electrical network balance sub-model and a mechanical balance sub-model of the motor.

[0130] Based on the above embodiment, optionally, the device also includes a parameter optimization module for constituting particles in the particle swarm algorithm based on the physical parameters of the current model; randomly initializing the particle position and particle velocity of each particle; iteratively updating the particle position and the particle velocity of each particle based on the particle swarm algorithm, and determining the updated fitness value based on the fitness function; when the iteration stop condition is met, taking the particle obtained by the current iteration as the optimal particle, and updating the current model based on the physical parameters constituting the optimal particle.

[0131] Based on the above embodiment, optionally, the parameter optimization module includes a fitness value updating unit, which is used to obtain the real current signal of the motor in a normal state and perform noise reduction processing on the real current signal; and determine the updated fitness value based on the output current signal of the current model and the real current signal after noise reduction processing.

[0132] Based on the above embodiment, optionally, the sample space construction module 330 is specifically used to extract features and reduce the dimension of the time domain signal of the fault sample data to obtain feature data corresponding to the fault sample data; and construct a sample space based on the feature data corresponding to the fault sample data.

[0133] Based on the above embodiment, optionally, the fault diagnosis model training module 340 includes an initial fault model construction module, which is used to establish a neural network model and initialize the weights and thresholds of the neural network model; the weights and thresholds of the neural network model are updated based on a preset optimization algorithm to obtain an initial fault diagnosis model.

[0134] The training device for the motor fault diagnosis model provided in the embodiment of the present invention can execute the training method for the motor fault diagnosis model provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0135] Embodiment 4

[0136] Figure 5 Schematic diagram of the structure of a motor fault diagnosis device provided by the fourth embodiment of the present invention. Figure 5 As shown, the device comprises:

[0137] The current signal acquisition module 410 is used to acquire the current signal of the motor to be diagnosed at the current moment and extract characteristic data of the current signal;

[0138] The fault diagnosis module 420 is used to input the characteristic data of the current signal into the fault diagnosis model corresponding to the motor to be diagnosed for diagnosis, so as to obtain the fault diagnosis result of the motor to be diagnosed; wherein, the fault diagnosis model is trained based on the training method of the motor fault diagnosis model described in any embodiment of the present invention.

[0139] The technical solution of this embodiment obtains the current signal of the motor to be diagnosed at the current moment, extracts the characteristic data of the current signal, and inputs the characteristic data of the current signal into the fault diagnosis model corresponding to the motor to be diagnosed for diagnosis, thereby obtaining the fault diagnosis result of the motor to be diagnosed, which can improve the accuracy of fault diagnosis.

[0140] The motor fault diagnosis device provided in the embodiment of the present invention can execute the motor fault diagnosis method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0141] Embodiment 5

[0142] Figure 6 1 is a schematic diagram of the structure of an electronic device provided in Embodiment 5 of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0143] like Figure 6 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0144] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0145] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 11 executes the various methods and processes described above, such as a training method for a fault diagnosis model of a motor, and / or a fault diagnosis method for a motor.

[0146] In some embodiments, the training method of the fault diagnosis model of the motor, and / or the fault diagnosis method of the motor may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, the training method of the fault diagnosis model of the motor, and / or one or more steps of the fault diagnosis method of the motor described above may be executed. Alternatively, in other embodiments, the processor 11 may be configured to execute the training method of the fault diagnosis model of the motor, and / or the fault diagnosis method of the motor by any other appropriate means (e.g., by means of firmware).

[0147] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0148] The computer program for implementing the training method of the fault diagnosis model of the motor of the present invention, and / or the fault diagnosis method of the motor can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or the block diagram are implemented. The computer program can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0149] Embodiment 6

[0150] Embodiment 6 of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to execute a training method for a motor fault diagnosis model, the method comprising:

[0151] Constructing a current model of the motor, wherein the current model includes multiple physical parameters inside the motor;

[0152] In response to a fault simulation operation on the current model, a parameter value of at least one physical parameter of the current model is changed, and the changed current model is run to obtain fault sample data corresponding to the fault simulation operation;

[0153] Constructing a sample space of the motor based on the fault sample data;

[0154] An initial fault diagnosis model is constructed, and the initial fault diagnosis model is trained based on the sample space to obtain a fault diagnosis model of the motor.

[0155] And / or, performing a motor fault diagnosis method, comprising:

[0156] Acquire the current signal of the motor to be diagnosed at the current moment, and extract characteristic data of the current signal;

[0157] The characteristic data of the current signal is input into the fault diagnosis model corresponding to the motor to be diagnosed to perform diagnosis, so as to obtain the fault diagnosis result of the motor to be diagnosed; wherein, the fault diagnosis model is trained based on the training method of the motor fault diagnosis model described in any embodiment of the present invention.

[0158] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0159] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0160] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0161] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0162] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0163] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for training a motor fault diagnosis model, characterized in that: include: Constructing a current model of the motor, wherein the current model includes multiple physical parameters inside the motor; In response to a fault simulation operation on the current model, a parameter value of at least one physical parameter of the current model is changed, and the changed current model is run to obtain fault sample data corresponding to the fault simulation operation; Constructing a sample space of the motor based on the fault sample data; An initial fault diagnosis model is constructed, and the initial fault diagnosis model is trained based on the sample space to obtain a fault diagnosis model of the motor.

2. The method according to claim 1, characterized in that The current model of the motor includes an electrical network balance sub-model and a mechanical balance sub-model of the motor.

3. The method according to claim 2, characterized in that After constructing the current model of the motor, the method further includes: The physical parameters of the current model constitute particles in a particle swarm algorithm; Randomly initializing the particle position and particle velocity of each of the particles; Iteratively updating the particle position and the particle velocity of each particle based on the particle swarm algorithm, and determining an updated fitness value based on a fitness function; When the iteration stop condition is met, the particle obtained in the current iteration is used as the optimal particle, and the current model is updated based on the physical parameters constituting the optimal particle.

4. The method according to claim 3, characterized in that The step of determining the updated fitness value based on the fitness function includes: Acquire a real current signal of the motor in a normal state, and perform noise reduction processing on the real current signal; An updated fitness value is determined based on the output current signal of the current model and the real current signal after noise reduction processing.

5. The method according to claim 1, characterized in that: The step of constructing a sample space of the motor based on the fault sample data comprises: Performing feature extraction and dimensionality reduction processing on the time domain signal of the fault sample data to obtain feature data corresponding to the fault sample data; A sample space is constructed based on the feature data corresponding to the fault sample data.

6. The method according to claim 1, characterized in that The constructing of the initial fault diagnosis model comprises: Establishing a neural network model and initializing the weights and thresholds of the neural network model; The weights and thresholds of the neural network model are updated based on a preset optimization algorithm to obtain an initial fault diagnosis model.

7. A method for diagnosing a motor fault, characterized in that: include: Acquire the current signal of the motor to be diagnosed at the current moment, and extract characteristic data of the current signal; The characteristic data of the current signal is input into the fault diagnosis model corresponding to the motor to be diagnosed to perform diagnosis, so as to obtain the fault diagnosis result of the motor to be diagnosed; wherein, the fault diagnosis model is trained based on the training method of the fault diagnosis model of the motor described in any one of claims 1-6.

8. A training device for a motor fault diagnosis model, characterized in that: include: A current model building module, used to build a current model of the motor, wherein the current model includes multiple physical parameters inside the motor; A fault simulation module, configured to change a parameter value of at least one physical parameter of the current model in response to a fault simulation operation on the current model, and to run the changed current model to obtain fault sample data corresponding to the fault simulation operation; A sample space construction module, used to construct a sample space of the motor based on the fault sample data; The fault diagnosis model training module is used to construct an initial fault diagnosis model, and train the initial fault diagnosis model based on the sample space to obtain the fault diagnosis model of the motor.

9. A motor fault diagnosis device, characterized in that: include: A current signal acquisition module, used to acquire the current signal of the motor to be diagnosed at the current moment, and extract characteristic data of the current signal; A fault diagnosis module is used to input the characteristic data of the current signal into the fault diagnosis model corresponding to the motor to be diagnosed for diagnosis, so as to obtain the fault diagnosis result of the motor to be diagnosed; wherein, the fault diagnosis model is trained based on the training method of the fault diagnosis model of the motor described in any one of claims 1-6.

10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the training method of the motor fault diagnosis model described in any one of claims 1-6, and / or the motor fault diagnosis method described in claim 7.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the training method of the motor fault diagnosis model described in any one of claims 1 to 6, and / or the motor fault diagnosis method described in claim 7 when executed.

Citation Information

Patent Citations

  • Spring operating mechanism circuit breaker modeling and fault diagnosis method

    CN109031114A

  • Numerical simulation drive motor turn-to-turn short circuit depth migration fault diagnosis method

    CN115510741A

  • Permanent magnet motor health state monitoring control method and system based on digital twinning

    CN117491869A

Cited By

  • Method and device for training network diagnosis model, equipment and storage medium

    CN120185999A

  • Training method of power system fault diagnosis model, and fault diagnosis method and device

    CN121388580A