Armored vehicle permanent magnet synchronous motor health monitoring model construction method and system based on reduced-order digital twin model

Abstract

The invention discloses an armored vehicle permanent magnet synchronous motor health monitoring model construction method and system based on a reduced-order digital twin model, and the method comprises the steps: building an electromagnetic-temperature coupling field finite element model based on the parameters of a permanent magnet synchronous motor, and simulating the operation parameters of the permanent magnet synchronous motor at a preset rotating speed; constructing an original feature set based on the operation parameters of the permanent magnet synchronous motor at the preset rotating speed; performing dimensionality reduction and cumulative variance interpretation on the original feature set by using a principal component analysis method to obtain a matrix formed by principal component feature vectors; an ordinary differential neural network is adopted to continuously solve an ordinary differential equation fitting historical state through dimension reduction time sequence data obtained through simulation, and an adjoint state of network parameters is reversely solved to form a permanent magnet synchronous motor digital twin reduced-order model; the coupling relation between the physical system and the preset initial value is relieved, the initial value problem is prevented from affecting the prediction effect, and reliable model and experience support is provided for health management operation and maintenance of the permanent magnet synchronous motor.

Description

Technical Field

[0001] The present invention belongs to the technical field of motor digital twins, and specifically relates to a method and system for constructing a health monitoring model of an armored vehicle permanent magnet synchronous motor based on a reduced-order digital twin model. Background Art

[0002] Permanent Magnet Synchronous Motor (PMSM) is a synchronous motor that uses permanent magnets as magnetic field sources. It has the advantages of low loss, high efficiency, high power density, good dynamic performance, low maintenance cost, small size, light weight, and reliable operation. It is widely used in industrial automation, electric vehicles, wind power generation and other fields.

[0003] The finite element method (FEM) is currently a common digital twin method for permanent magnet synchronous motors to simulate the dynamic performance of permanent magnet synchronous motors under multi-physical fields. Although the finite element method can provide high-precision electromagnetic field and temperature field distribution, it takes a long time to calculate, occupies a lot of memory, and has high requirements on computing resources. This disadvantage is particularly prominent in the process of transient or parametric research, and cannot meet the real-time state interaction and online state monitoring of the digital twin model of the permanent magnet synchronous motor and the physical entity.

[0004] In order to meet the basic requirements of virtual-real interaction between digital twin models and physical entities, it is necessary to reduce the degrees of freedom in the finite element model and reduce the dimensionality of the high-dimensional data generated by the finite element model simulation. There are currently a variety of reduced order models (ROM) technologies. Among them, principal component analysis (PCA) is an effective order reduction algorithm. PCA is used to extract the main dynamic features from high-dimensional data such as current, air gap magnetic flux, and electromagnetic torque obtained from finite element simulation, and reduce the high-dimensional system to a smaller feature space.

[0005] Neural Ordinary Differential Equations (Neural ODEs) is a deep learning model that uses solving ordinary differential equations as the main calculation method. It can decouple the boundary conditions and initial conditions of the physical system from the system state prediction process. It is suitable for processing continuous time series data and is suitable as a proxy model for the digital twin model of permanent magnet synchronous motors. The health state of permanent magnet synchronous motors is characterized by state quantities, such as current time-frequency domain characteristics, torque pulsation, vibration, temperature, etc. These parameters are nonlinear and dynamically changing time series. Neural ODEs have excellent continuous modeling capabilities and higher parameter efficiency than traditional layer neural networks. It can flexibly capture the continuous dynamic characteristics of physical systems.

[0006] In the prior art, a Chinese patent with publication number CN117491869A provides a permanent magnet synchronous motor health status monitoring and control method and system based on digital twins, proposes a current prediction model based on a four-section Runge-Kutta solver, uses a particle swarm optimization algorithm to optimize the motor parameters, and inputs the optimal parameters into the zero-beat prediction motor model to achieve motor control. This solution does not predict the fault state and operating temperature of the permanent magnet synchronous motor; a Chinese patent with publication number CN114492137A provides a motor status monitoring and prediction method based on finite elements and digital twins. , a finite element coupled circuit combination mathematical model for motor state prediction was constructed. The overall technology is based on the finite element coupled circuit combination model, and the model is not reduced in order, which consumes a lot of computing resources; the Chinese patent with publication number CN118535914A provides a key equipment fault diagnosis method for power supply system based on digital twins, which constructs a fault behavior twin model for key equipment in the power supply system, generates a fault data set, and constructs and updates a fault state perception twin model based on the fault data set and physical information neural network, and uses the updated fault state perception twin model to perform fault diagnosis on the sample to be tested. This solution focuses on the diagnosis and classification of faults that have occurred, and does not perform work on fault degree prediction. Summary of the invention

[0007] In order to solve the problems existing in the prior art, the present invention provides a permanent magnet synchronous motor health monitoring method based on a reduced-order digital twin model. First, a physical model of the permanent magnet synchronous motor is established according to relevant parameters of the target motor, and multi-physical field simulation of the permanent magnet synchronous motor is performed using the finite element method to generate an original data set. Principal component analysis is performed on the data set, and the number of principal components is determined according to the cumulative explained variance method to obtain a low-dimensional matrix. The reduced-dimensional data is used to train an ordinary differential neural network to track the health status of the permanent magnet synchronous motor stator winding, the degradation state of the permanent magnet and the real-time temperature of the motor, so as to finally realize the health status monitoring of the permanent magnet synchronous motor.

[0008] In order to achieve the above-mentioned object, in a first aspect, the present invention provides a method for constructing a health monitoring model of a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model, comprising the following steps: establishing an electromagnetic-temperature coupling field finite element model based on the parameters of the permanent magnet synchronous motor, and using the electromagnetic-temperature coupling field finite element model to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed;

[0009] Constructing an original feature set based on the operating parameters of the permanent magnet synchronous motor at a preset speed;

[0010] The principal component analysis method is used to reduce the dimension of the original feature set and accumulate the explained variance to obtain a matrix consisting of the principal component eigenvectors;

[0011] An ordinary differential neural network is used to continuously solve the ordinary differential equations to fit the historical state of the reduced-dimensional time series data obtained by simulation, and the adjoint state of the network parameters is solved in reverse to form a digital twin reduced-order model of the permanent magnet synchronous motor.

[0012] Furthermore, an electromagnetic-temperature coupling field finite element model is established based on the parameters of the permanent magnet synchronous motor, and the electromagnetic-temperature coupling field finite element model is used to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed, including:

[0013] According to the design parameter information of the target permanent magnet synchronous motor, a two-dimensional model of the permanent magnet synchronous motor is established in COMSOL software;

[0014] Use the built-in AC / DC module of COMSOL software to perform electromagnetic field analysis on the motor, set the motor rotor parameters, stator parameters and cylindrical coordinates, divide the stator and rotor boundaries, define vector changes, and set the material properties of each part of the motor;

[0015] Define boundaries and finite element analysis constraints for different physical fields, determine coupling relationships, create meshes and solve them to obtain three-phase currents, back electromotive force, electromagnetic torque, motor stator temperature, and motor rotor temperature at preset speeds.

[0016] Furthermore, when creating the grid and solving the problem, the grid is automatically created by COMSOL, the grid density is determined according to the physical field settings, and then the grid structure is manually adjusted, the automatically created grid sequence is deleted, and a boundary layer grid is set at the boundary between the stator and the rotor.

[0017] Furthermore, the original feature set constructed based on the operating parameters of the permanent magnet synchronous motor at a preset speed includes:

[0018] The electromagnetic-temperature coupled field finite element model is used to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed, and the operating parameters include three-phase current, three-phase back electromotive force, electromagnetic torque, motor stator temperature, and motor rotor temperature at the preset speed; the negative-sequence current component of the simulated three-phase current signal and the three-phase back electromotive force, the output electromagnetic torque signal, the motor stator temperature, and the motor rotor temperature together constitute an original feature set, the asymmetry degree of the negative-sequence component of the three-phase current signal is used to characterize the degree of short circuit between the turns of the motor stator winding, the amplitude of the three-phase back electromotive force is used to characterize the degree of demagnetization of the motor permanent magnet, and the electromagnetic torque characterizes the health status of the motor.

[0019] Furthermore, the principal component analysis method is used to reduce the dimension of the original feature set and accumulate the explained variance, and the matrix composed of the principal component eigenvectors is obtained, including:

[0020] The time series of the operating parameters obtained by simulation are organized into a matrix form, where each column represents a feature and each row represents a time point;

[0021] Each feature is processed by returning the mean to zero and normalizing the variance to obtain the standardized matrix X';

[0022] Calculate the covariance matrix of the standardized matrix X', perform eigenvalue decomposition on the covariance matrix C, and obtain eigenvalues ​​and eigenvectors;

[0023] According to the different monitoring targets of permanent magnet synchronous motors, the cumulative explained variance method is used to select the appropriate number of principal components for the monitoring targets, calculate the variance contribution rate corresponding to each principal component, accumulate the variance contribution rates, and select several features with the largest contribution as the principal components of the corresponding monitoring targets; retain the eigenvectors corresponding to the principal components, project the original data onto the principal components, and obtain the matrix composed of the eigenvectors of the principal components.

[0024] Furthermore, an ordinary differential neural network is used to continuously solve the ordinary differential equation to fit the historical state of the reduced-dimensional time series data obtained by simulation, and reversely solve the adjoint state of the network parameters to form a permanent magnet synchronous motor digital twin reduced-order model including:

[0025] Subdivide the number of network layers of the ordinary differential neural network to obtain its differential form, and solve the hidden state of each layer according to the initial state;

[0026] According to different monitoring targets, the reduced-dimensional data is organized into a form suitable for time series modeling, and the input data is constructed according to the time window;

[0027] The input data is used to train the ordinary differential neural network, and the measured data of the permanent magnet synchronous motor is fitted by solving the ordinary differential equation to complete the solution of the state quantity at the termination time, and then the loss function of the forward calculation is solved, and the model parameters of the ordinary differential neural network are updated by the adjoint state method to obtain the adjoint state of the hidden state; similarly, the adjoint state of the ordinary differential neural network parameters and time is obtained, and the global parameter gradient is obtained by reverse iterative solution, and the ordinary differential neural network parameters are updated to form a permanent magnet synchronous motor digital twin reduced-order model.

[0028] In the second aspect, the present invention provides a permanent magnet synchronous motor health monitoring method based on a reduced-order digital twin model, and the permanent magnet synchronous motor digital twin reduced-order model obtained based on the model construction method specifically includes: obtaining the three-phase current, three-phase back electromotive force, electromagnetic torque, and temperature data of the permanent magnet synchronous motor, and inputting the ordinary differential neural network model obtained by reducing the order of the multi-physical field coupling finite element model of the permanent magnet synchronous motor, that is, in the digital twin reduced-order model of the permanent magnet synchronous motor, monitoring the health status within a set time to determine whether there is a potential fault possibility; setting an error threshold, when the error between the current or back electromotive force and the prediction of the ordinary differential neural network model exceeds the threshold, warning of stator winding inter-turn short-circuit fault, permanent magnet demagnetization fault and overheating, and displaying the degree of short-circuit fault, permanent magnet demagnetization rate and overheating according to the error size, and further forming an adjustment strategy.

[0029] On the other hand, the present invention provides a method for constructing a health monitoring model of a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model, comprising a finite element simulation module, a feature set acquisition module, a feature set processing module, and a fitting solution module;

[0030] The finite element simulation module establishes an electromagnetic-temperature coupling field finite element model based on the parameters of the permanent magnet synchronous motor, and uses the electromagnetic-temperature coupling field finite element model to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed;

[0031] The feature set acquisition module constructs an original feature set based on the operating parameters of the permanent magnet synchronous motor at a preset speed;

[0032] The feature set processing module uses the principal component analysis method to reduce the dimension of the original feature set and accumulate the explained variance to obtain a matrix composed of the principal component eigenvectors;

[0033] The fitting and solving module uses an ordinary differential neural network to continuously solve the ordinary differential equations to fit the historical state of the reduced-dimensional time series data obtained by simulation, and reversely solves the accompanying state of the network parameters to form a reduced-order digital twin model of the permanent magnet synchronous motor.

[0034] The present invention can also provide a computer device, including a processor and a memory, the memory is used to store a computer executable program, the processor reads the computer executable program from the memory and executes it, and when the processor executes the computer executable program, it can implement the armored vehicle permanent magnet synchronous motor health monitoring model construction method based on the reduced-order digital twin model and the permanent magnet synchronous motor health monitoring method based on the reduced-order digital twin model described in the present invention.

[0035] At the same time, a computer-readable storage medium is provided, in which a computer program is stored. When the computer program is executed by a processor, the method for constructing a health monitoring model of a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model and the method for health monitoring of a permanent magnet synchronous motor based on a reduced-order digital twin model described in the present invention can be implemented.

[0036] Compared with the prior art, the present invention has at least the following beneficial effects: data dimension reduction is performed on the established refined permanent magnet synchronous motor multi-physical field coupled finite element model, and an ordinary differential neural network is trained through the dimension reduction simulation data as a digital twin reduced-order model of the permanent magnet synchronous motor. The coupling relationship between the physical system and the preset initial value is released to avoid the initial value problem affecting the prediction effect. The digital twin reduced-order model based on the ordinary differential neural network is used to predict the degree of stator inter-turn short circuit, monitor the degree of permanent magnet demagnetization, detect temperature and give over-temperature warning during the operation of the permanent magnet synchronous motor, providing a reliable model and experience support for the health management and operation of the permanent magnet synchronous motor. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a technical flow chart of a permanent magnet synchronous motor health monitoring method based on digital twin;

[0038] Figure 2 Flowchart for finite element modeling;

[0039] Figure 3 This is a schematic diagram of the cumulative interpretation method;

[0040] Figure 4 Schematic diagram of the permanent magnet synchronous health monitoring process. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are 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 are within the scope of protection of the present invention.

[0042] The performance of permanent magnet synchronous motor is affected by two main factors: electromagnetic and temperature. The present invention can more accurately simulate the behavior of the motor under different working conditions, especially the operating parameters at a preset speed, by establishing a finite element model that couples these two factors, which can help understand the complex interactions inside the motor, and construct an original feature set through the simulated operating parameters, providing basic data for subsequent analysis and modeling; these original features contain key information about the operation of the motor, and the principal component analysis method is used to reduce the dimension and accumulate the explained variance, which helps to screen out the key features that have the greatest impact on the performance of the motor, which can not only reduce the computational complexity, but also improve the prediction accuracy of the model, and the model can focus more on the most representative features; the ordinary differential neural network enables the model to capture the dynamic characteristics of the data by continuously solving ordinary differential equations, thereby more accurately simulating the actual operation of the motor, allowing the model to learn the law of motor state changes over time, and then can perform advanced applications such as state prediction and fault diagnosis; the digital twin reduced-order model finally formed combines the physical characteristics of the electromagnetic-temperature coupling field and the learning results of the motor operation data, which can reduce the complexity of the model while ensuring the prediction accuracy, and is more suitable for application scenarios such as real-time monitoring and predictive maintenance, which helps to discover potential problems in advance and intervene in time, thereby helping to improve the reliability and service life of the motor.

[0043] According to the design parameter information of the target permanent magnet synchronous motor, a two-dimensional model of the permanent magnet synchronous motor is established in COMSOL software. The change of the electromagnetic field inside the motor along the axial direction is ignored. The two-dimensional model of the permanent magnet synchronous motor is used as the analysis object to effectively analyze the running state of the motor. The finite element modeling process is as follows: Figure 2 shown.

[0044] The built-in AC / DC module of COMSOL software is selected to perform electromagnetic field analysis on the motor. The transient solution of the AC / DC module is used to obtain the timing information of the two-dimensional electromagnetic field when the motor is running, and to simulate multiple electromagnetic field quantities that are inconvenient to measure (magnetic field strength of the air gap, loss density in the magnet, etc.). First, a steady-state study is performed on the motor, and then a transient study is performed on this basis.

[0045] First, define the parameters, including defining variables, setting functions, and setting material parameters. Define the number of poles, rotor diameter, and magnet height of the motor rotor, define the number of stator slots, slot winding type, define cylindrical coordinates for subsequent definition of motor speed, set stator and rotor boundary division, and define vector changes for subsequent viewing of state values ​​in a certain direction. Define the material properties of each part of the motor, add Air, Soft Iron (Without Losses), N50, Aluminum, Structural steel, and Copper materials to the model tree, then set the corresponding materials for the air gap, permanent magnet, coil, and core, and set relevant parameters such as conductivity, relative dielectric constant, magnetic permeability, magnetic coercivity, loss coefficient, and specific heat capacity to complete the material settings.

[0046] After completing the model import and parameter definition, perform geometry settings and physical field settings. First, perform electromagnetic field settings, apply electromagnetic field finite element analysis constraints, set magnetic flux conservation conditions for air and iron core; define the magnetization direction and intensity of permanent magnets; set Ampere's law conditions for the stator iron core; determine the number of stator winding coil turns based on motor parameters, define current density and voltage excitation for the winding, set loss calculations for coils and permanent magnets; perform global torque calculations; and define motor speed.

[0047] After completing the electromagnetic field setting, the temperature field setting is performed. Set material parameters such as thermal conductivity, specific heat capacity, and density in the heat conduction module; couple the electromagnetic loss as a heat source to the heat conduction module, including the copper loss of the winding, the iron loss of the core, and the permanent magnet loss; set the natural convection and radiation boundary conditions for the model to simulate the actual operating conditions of the motor.

[0048] After completing the geometric model and physical field settings, meshing is performed. First, the mesh is automatically created through COMSOL, and the mesh density is determined according to the physical field settings. Then the mesh structure is manually adjusted, the automatically created mesh sequence is deleted, and the boundary layer mesh is set at the boundary between the stator and the rotor. The boundary mesh density is increased to calculate the electromagnetic effect in the air gap at the junction of the stator and the rotor. After completing the meshing, the solver is set to solve the state parameters and simulate the three-phase current I of the motor at the preset speed. a ,I b ,I c 、Three-phase back electromotive force e a 、e b 、e c , speed n, electromagnetic torque T e , motor stator temperature Temp1 and motor rotor temperature Temp2 state quantities.

[0049] The operating state of the permanent magnet synchronous motor can be characterized by the state quantity values ​​obtained by measurement and simulation as well as the time and frequency domain characteristics. The negative-sequence current component of the three-phase current signal obtained by simulation, together with the three-phase back electromotive force, the output electromagnetic torque signal, the motor stator temperature, and the motor rotor temperature constitute the original feature set, which provides original data support for the next step of data dimensionality reduction using the principal component analysis method. The asymmetry degree of the negative-sequence component of the three-phase current signal is mainly used to characterize the degree of short circuit between the turns of the motor stator winding, the amplitude of the three-phase back electromotive force is mainly used to characterize the degree of demagnetization of the motor permanent magnet, and the output electromagnetic torque mainly characterizes the health status of the motor from the side.

[0050] The time series of different physical quantities obtained by simulation are organized into a matrix form for principal component analysis. All data are organized into a matrix X, where each column represents a feature and each row represents a time point. Assuming that the number of samples is N and the number of features is M = 9 (three-phase current negative sequence component, three-phase back electromotive force, electromagnetic torque, stator temperature, rotor temperature), then:

[0051] First, the data is standardized, and the mean and variance of each feature are normalized according to the following formula:

[0052]

[0053] Where μ is the mean of the feature and σ is the standard deviation of the feature.

[0054] The mean of each column of the standardized data matrix X' is 0 and the variance is 1.

[0055] After standardization, calculate the covariance matrix C of the standardized matrix X':

[0056]

[0057] Each element of the covariance matrix represents the correlation between features.

[0058] Perform eigenvalue decomposition on the covariance matrix C to obtain eigenvalues ​​and eigenvectors:

[0059] C.v i =λ i ·v i

[0060] where λ i is the eigenvalue, which indicates the variance explained by each principal component; v i is the eigenvector, indicating the direction of each principal component.

[0061] For different monitoring targets of permanent magnet synchronous motors (winding short circuit degree, permanent magnet demagnetization degree, real-time temperature curve), the cumulative explained variance method is used to select the appropriate number of principal components for the monitoring targets, normalize the eigenvalues, calculate the variance contribution rate corresponding to each principal component, and accumulate the variance contribution rates. Usually, the number of principal components with a cumulative explained variance of 85% is selected. The schematic diagram of the cumulative explained variance method is shown in the figure below: Figure 3 As shown, there are 9 candidate features. By calculating the contribution of each feature, it can be seen from the figure that the cumulative contribution of the first four features exceeds 85%, so the first four features are selected as the main components of a monitoring target.

[0062] Retain the eigenvectors corresponding to the principal components, project the original data onto these principal components, and obtain the reduced-dimensional data set. The reduced-dimensional data matrix X is calculated by the following formula: reduced :

[0063] X reduced =X'·P

[0064] Where P is a matrix consisting of the eigenvectors of the first four principal components.

[0065] The ordinary differential neural network continuously solves the ordinary differential equation to fit the historical state of the reduced-dimensional time series data obtained by simulation, reversely solves the adjoint state of the network parameter θ, and finally obtains the global gradient to achieve the update of the ordinary differential neural network parameters. Then, the state of the permanent magnet synchronous motor is predicted through the trained ordinary differential neural network to realize health status monitoring.

[0066] The number of network layers of an ordinary differential neural network can be regarded as a continuous variable, which can be infinitely subdivided to obtain its differential form:

[0067]

[0068] Among them, x t is the state quantity of the permanent magnet synchronous motor; θ is the parameter to be learned in the network; t is the time.

[0069] By setting the initial states x0, t0, t1 in advance, we can solve the hidden states of each layer:

[0070]

[0071] First, for different monitoring targets, the reduced-dimensional data needs to be organized into a form suitable for time series modeling, and the input data is constructed according to the time window. Assuming the window size is T, the input can be expressed as:

[0072] X input =[X(tT),X(t-T+1),...,X(t)]

[0073] The output is:

[0074] Y output =X(t+1)

[0075] Where X(t) is the dimension-reduced data.

[0076] The input data is used to train the ordinary differential neural network, and the measured data of the permanent magnet synchronous motor is fitted by solving the ordinary differential equation. After the state quantity at the termination time is solved, the loss function L of the forward calculation is solved:

[0077]

[0078] The adjoint state method is used to update the model parameters of the ordinary differential neural network, where the adjoint state is defined as:

[0079]

[0080] Where a(t) is the adjoint state of the hidden state x under variable t.

[0081] Similarly, the adjoint state of the ordinary differential neural network parameter θ and time t under variable t can be expressed as:

[0082]

[0083] Since the state quantity at the termination time is only related to the constant differential neural network parameter state θ at the previous time n-1 Therefore, we set a θ (t) = 0, reverse iterate and solve a according to a(t) θ (t) and a t (t), and obtain the parameter gradient for solving the global problem:

[0084]

[0085] The ordinary differential neural network parameter θ is updated to form a reduced-order model of the permanent magnet synchronous motor digital twin, and the updated reduced-order model is used to predict the change trend of the health status of the permanent magnet synchronous motor in a subsequent time period.

[0086] The health monitoring process diagram is as follows: Figure 4The three-phase current, three-phase back EMF, electromagnetic torque and temperature data of the permanent magnet synchronous motor are obtained through test bench testing, and input into the ordinary differential neural network model obtained by reducing the order of the multi-physics field coupling finite element model of the permanent magnet synchronous motor to monitor the short circuit degree, permanent magnet demagnetization degree and real-time temperature of the permanent magnet synchronous motor. By analyzing the health status over a period of time, it is judged whether there is a potential fault possibility; the error threshold is set. When the error between the current or back EMF and the prediction of the ordinary differential neural network model exceeds the threshold, the stator winding inter-turn short circuit fault, permanent magnet demagnetization fault and overheating phenomenon are warned, and the short circuit fault degree, permanent magnet demagnetization rate and overheating degree are displayed according to the error size, so as to further form relevant adjustment strategies.

[0087] Based on the concept of the method, the present invention also provides a method for constructing a health monitoring model of a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model, comprising a finite element simulation module, a feature set acquisition module, a feature set processing module, and a fitting solution module;

[0088] The finite element simulation module establishes an electromagnetic-temperature coupling field finite element model based on the parameters of the permanent magnet synchronous motor, and uses the electromagnetic-temperature coupling field finite element model to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed;

[0089] The feature set acquisition module constructs an original feature set based on the operating parameters of the permanent magnet synchronous motor at a preset speed;

[0090] The feature set processing module uses the principal component analysis method to reduce the dimension of the original feature set and accumulate the explained variance to obtain a matrix composed of the principal component eigenvectors;

[0091] The fitting and solving module uses an ordinary differential neural network to continuously solve the ordinary differential equations to fit the historical state of the reduced-dimensional time series data obtained by simulation, and reversely solves the accompanying state of the network parameters to form a reduced-order digital twin model of the permanent magnet synchronous motor.

[0092] On the other hand, the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it can implement the method for constructing a health monitoring model of an armored vehicle permanent magnet synchronous motor based on a reduced-order digital twin model described in the present invention, and can also implement a permanent magnet synchronous motor health monitoring method based on a reduced-order digital twin model.

[0093] The present invention can also provide a computer device, including a processor and a memory, the memory is used to store a computer executable program, the processor reads the computer executable program from the memory and executes it, and when the processor executes the computer executable program, it can implement the method for constructing a health monitoring model of a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model described in the present invention, and can also implement a permanent magnet synchronous motor health monitoring method based on a reduced-order digital twin model.

[0094] The computer device may be a laptop computer, a desktop computer or a workstation.

[0095] The processor may be a central processing unit (CPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), or an off-the-shelf field programmable gate array (FPGA).

[0096] The memory of the present invention may be an internal storage unit of a laptop computer, a desktop computer or a workstation, such as a memory or a hard disk; or an external storage unit, such as a mobile hard disk or a flash memory card.

[0097] Computer-readable storage media may include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented by any method or technology for storing information such as computer-readable instructions, data structures, program modules or other data. Computer-readable storage media may include: read-only memory (ROM), random access memory (RAM), solid-state drive (SSD) or optical disk, etc. Among them, random access memory may include resistance random access memory (ReRAM) and dynamic random access memory (DRAM).

[0098] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A method for constructing a health monitoring model for a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model, characterized in that: The following steps are involved: An electromagnetic-temperature coupling field finite element model is established based on the parameters of the permanent magnet synchronous motor, and the electromagnetic-temperature coupling field finite element model is used to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed; Constructing an original feature set based on the operating parameters of the permanent magnet synchronous motor at a preset speed; The principal component analysis method is used to reduce the dimension of the original feature set and accumulate the explained variance to obtain a matrix consisting of the principal component eigenvectors; An ordinary differential neural network is used to continuously solve the ordinary differential equations to fit the historical state of the reduced-dimensional time series data obtained by simulation, and the adjoint state of the network parameters is solved in reverse to form a digital twin reduced-order model of the permanent magnet synchronous motor.

2. The method for constructing a health monitoring model for an armored vehicle permanent magnet synchronous motor based on a reduced-order digital twin model according to claim 1 is characterized in that: An electromagnetic-temperature coupling field finite element model is established based on the parameters of the permanent magnet synchronous motor. The electromagnetic-temperature coupling field finite element model is used to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed, including: According to the design parameter information of the target permanent magnet synchronous motor, a two-dimensional model of the permanent magnet synchronous motor is established in COMSOL software; Use the built-in AC / DC module of COMSOL software to perform electromagnetic field analysis on the motor, set the motor rotor parameters, stator parameters and cylindrical coordinates, divide the stator and rotor boundaries, define vector changes, and set the material properties of each part of the motor; Define boundaries and finite element analysis constraints for different physical fields, determine coupling relationships, create meshes and solve them to obtain three-phase currents, back electromotive force, electromagnetic torque, motor stator temperature, and motor rotor temperature at preset speeds.

3. The method for constructing a health monitoring model for a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model according to claim 2 is characterized in that: When creating the grid and solving the problem, the grid is automatically created by COMSOL, the grid density is determined according to the physical field setting, and then the grid structure is manually adjusted, the automatically created grid sequence is deleted, and the boundary layer grid is set at the boundary between the stator and the rotor.

4. The method for constructing a health monitoring model for an armored vehicle permanent magnet synchronous motor based on a reduced-order digital twin model according to claim 1, characterized in that: The original feature set constructed based on the operating parameters of the permanent magnet synchronous motor at a preset speed includes: The electromagnetic-temperature coupled field finite element model is used to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed, and the operating parameters include three-phase current, three-phase back electromotive force, electromagnetic torque, motor stator temperature, and motor rotor temperature at the preset speed; the negative-sequence current component of the simulated three-phase current signal and the three-phase back electromotive force, the output electromagnetic torque signal, the motor stator temperature, and the motor rotor temperature together constitute an original feature set, the asymmetry degree of the negative-sequence component of the three-phase current signal is used to characterize the degree of short circuit between the turns of the motor stator winding, the amplitude of the three-phase back electromotive force is used to characterize the degree of demagnetization of the motor permanent magnet, and the electromagnetic torque characterizes the health status of the motor.

5. The method for constructing a health monitoring model for a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model according to claim 1, characterized in that: The principal component analysis method is used to reduce the dimension of the original feature set and accumulate the explained variance, and the matrix composed of the principal component eigenvectors is obtained, including: The time series of the operating parameters obtained by simulation are organized into a matrix form, where each column represents a feature and each row represents a time point; Each feature is processed by returning the mean to zero and normalizing the variance to obtain the standardized matrix X'; Calculate the covariance matrix of the standardized matrix X', perform eigenvalue decomposition on the covariance matrix C, and obtain eigenvalues ​​and eigenvectors; According to the different monitoring targets of permanent magnet synchronous motors, the cumulative explained variance method is used to select the appropriate number of principal components for the monitoring targets, calculate the variance contribution rate corresponding to each principal component, accumulate the variance contribution rates, and select several features with the largest contribution as the principal components of the corresponding monitoring targets; retain the eigenvectors corresponding to the principal components, project the original data onto the principal components, and obtain the matrix composed of the eigenvectors of the principal components.

6. The method for constructing a health monitoring model for a permanent magnet synchronous motor of an armored vehicle based on a reduced-order digital twin model according to claim 1, characterized in that: The ordinary differential neural network is used to continuously solve the ordinary differential equation to fit the historical state of the reduced-dimensional time series data obtained by simulation, and reversely solve the adjoint state of the network parameters to form a permanent magnet synchronous motor digital twin reduced-order model including: Subdivide the number of network layers of the ordinary differential neural network to obtain its differential form, and solve the hidden state of each layer according to the initial state; According to different monitoring targets, the reduced-dimensional data is organized into a form suitable for time series modeling, and the input data is constructed according to the time window; The input data is used to train the ordinary differential neural network, and the measured data of the permanent magnet synchronous motor is fitted by solving the ordinary differential equation to complete the solution of the state quantity at the termination time, and then the loss function of the forward calculation is solved, and the model parameters of the ordinary differential neural network are updated by the adjoint state method to obtain the adjoint state of the hidden state; similarly, the adjoint state of the ordinary differential neural network parameters and time is obtained, and the global parameter gradient is obtained by reverse iterative solution, and the ordinary differential neural network parameters are updated to form a permanent magnet synchronous motor digital twin reduced-order model.

7. A permanent magnet synchronous motor health monitoring method based on a reduced-order digital twin model, characterized in that: Based on the permanent magnet synchronous motor digital twin reduced-order model obtained by the model building method described in any one of claims 1 to 6, the three-phase current, three-phase back electromotive force, electromagnetic torque, and temperature data of the permanent magnet synchronous motor are obtained, and the ordinary differential neural network model obtained by reducing the order of the multi-physical field coupled finite element model of the permanent magnet synchronous motor is input, that is, in the permanent magnet synchronous motor digital twin reduced-order model, the health status within a set time is monitored to determine whether there is a possibility of potential failure; An error threshold is set. When the error between the current or back-electromotive force and the prediction of the ordinary differential neural network model exceeds the threshold, an early warning of stator winding inter-turn short-circuit fault, permanent magnet demagnetization fault and overheating is issued. The degree of short-circuit fault, permanent magnet demagnetization rate and overheating are displayed according to the error size, and an adjustment strategy is further formed.

8. A system for building a health monitoring model for permanent magnet synchronous motors of armored vehicles based on a reduced-order digital twin model, characterized in that: It includes finite element simulation module, feature set acquisition module, feature set processing module and fitting solution module; The finite element simulation module establishes an electromagnetic-temperature coupling field finite element model based on the parameters of the permanent magnet synchronous motor, and uses the electromagnetic-temperature coupling field finite element model to simulate the operating parameters of the permanent magnet synchronous motor at a preset speed; The feature set acquisition module constructs an original feature set based on the operating parameters of the permanent magnet synchronous motor at a preset speed; The feature set processing module uses the principal component analysis method to reduce the dimension of the original feature set and accumulate the explained variance to obtain a matrix composed of the principal component eigenvectors; The fitting and solving module uses an ordinary differential neural network to continuously solve the ordinary differential equations to fit the historical state of the reduced-dimensional time series data obtained by simulation, and reversely solves the accompanying state of the network parameters to form a reduced-order digital twin model of the permanent magnet synchronous motor.

9. A computer device, characterized in that: It includes a processor and a memory, the memory is used to store computer executable programs, the processor reads part or all of the computer executable programs from the memory and executes them, and when the processor executes part or all of the computer executable programs, it can implement the armored vehicle permanent magnet synchronous motor health monitoring model construction method based on the reduced-order digital twin model or the permanent magnet synchronous motor health monitoring method based on the reduced-order digital twin model as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: A computer program is stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the method for constructing a health monitoring model of an armored vehicle permanent magnet synchronous motor based on a reduced-order digital twin model or the permanent magnet synchronous motor health monitoring method based on a reduced-order digital twin model as described in any one of claims 1 to 7.

Images