Motor condition monitoring and prediction method based on finite element and digital twin

By building a finite element coupled circuit combination model and a digital twin hardware system, the problem of motor status prediction is solved, real-time monitoring and fault prediction of motor status are realized, and economic and safety risks brought about by motor failure are reduced.

CN114492137BActive Publication Date: 2024-11-15CHINA TOBACCO HENAN IND CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210118272.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-02-08
Publication Date
2024-11-15
Estimated Expiration
2042-02-08

AI Technical Summary

Technical Problem

The existing technology is difficult to effectively predict the motor status, resulting in failure and shutdown, causing economic losses and safety hazards, and lacks reliable early fault monitoring methods.

Method used

Using the motor state monitoring and prediction method based on finite element and digital twins, the finite element coupling circuit combination model of winding rotor induction motor is constructed, and the hardware system is built, and the digital twin processing unit and simulation platform are used for real-time monitoring and prediction, considering the motor The special geometric shape and winding structure are added to search coils and eccentricity to achieve real-time fault prediction.

Benefits of technology

Real-time monitoring and fault prediction of motor status are realized, economic losses and safety accidents caused by motor failure are reduced, and new detection adaptability and prediction methods are provided.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114492137B_ABST
    Figure CN114492137B_ABST
Patent Text Reader

Abstract

The present invention discloses a motor state monitoring and prediction method based on finite element and digital twin. The main design concept of the present invention is to construct a finite element coupled circuit combination mathematical model for motor state prediction without considering factors such as magnetic saturation, iron loss and capacitance effect in the motor model, and build a corresponding hardware system. The simulation is realized through the aforementioned model and hardware system, and it is used to monitor and predict the motor state. Specifically, the mathematical model takes into account the special geometric shape and winding structure of the motor, and can be flexibly adjusted through convenient matrix transformation, and any number of search coils and eccentricity and other faults can be added therein, so it is particularly suitable for real-time fault prediction of winding rotor induction motor. The present invention has strong detection adaptability, can greatly reduce economic losses and safety accidents caused by motor failures, and opens up a new way for real-time monitoring of the state of the motor and fault prediction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of equipment fault diagnosis, and in particular to a motor state monitoring and prediction method based on finite element and digital twin. Background Art

[0002] In recent years, with the rapid development of power electronics technology and the promotion of new energy technology, motors have played an irreplaceable role in the power system and power system of modern industry. As the core energy and power equipment of industrial production systems, the operating reliability of motors is of great significance to the stable operation of the entire system. Once the motor stops due to some serious fault, it will cause huge economic losses and even cause safety accidents, endangering the life safety of operators. Therefore, it is particularly important to provide timely warning and reliable monitoring of early faults of motors. Summary of the invention

[0003] In view of the above, the present invention aims to provide a motor state monitoring and prediction method based on finite element and digital twin to solve the problem of how to reliably predict the motor state.

[0004] The technical solution adopted by the present invention is as follows:

[0005] A motor condition monitoring and prediction method based on finite element and digital twin, including:

[0006] Step S1, constructing a finite element coupled circuit combination model of a winding rotor induction motor, specifically including the following: a model construction link, a circuit identification link, and a finite element calculation link of an inductance matrix;

[0007] Step S2, building a hardware system for motor state monitoring and prediction based on finite element and digital twin, specifically including: connecting the digital twin processing unit with the simulation platform data for real-time simulation, and inputting the sampled signals of the motor voltage and current acquisition circuit into the digital twin processing unit, and detecting the rotor position signal through the angular displacement sensor and inputting it into the digital twin processing unit; wherein the digital twin processing unit is configured as a finite element coupled circuit combination model of the winding rotor induction motor, and is used to output the prediction result;

[0008] Step S3, using the simulation platform to perform real-time processing to monitor and predict the motor state, specifically including the following: programming link, position tracking link, model implementation link and time step selection link.

[0009] In at least one possible implementation, the model building step includes:

[0010] Obtaining a total magnetic flux received by any motor circuit from other motor circuits, specifically, the total magnetic flux is equal to the sum of the magnetic fluxes received from other motor circuits; and based on a given motor circuit, the magnetic flux received from another motor circuit is related to the current flowing in the transmitter and the magnetic coupling between the two circuits; wherein the magnetic coupling is a function of the rotor angular displacement;

[0011] A dynamic model is obtained by using a single matrix equation to cover the above relationship between magnetic flux and current, and combining it with the voltage generated by the change of magnetic flux of any motor circuit; wherein the matrix equation includes the pre-obtained inductance matrix function;

[0012] The electromagnetic torque is calculated using the dynamic model, and the back electromotive force is obtained based on a first matrix that predefines the coupling relationship between the search coil and other motor circuits.

[0013] In at least one possible implementation, the circuit identification step includes:

[0014] Three winding coils are arranged on the motor stator and the motor rotor respectively, and a search coil is arranged around the stator teeth, so that the seven coil circuits constitute the hardware architecture of the finite element coupled circuit combination model of the winding rotor induction motor.

[0015] In at least one possible implementation, the finite element calculation step of the inductance matrix includes:

[0016] The inductance matrix functions of multiple rotor positions are calculated based on two-dimensional finite elements, and a lookup table is established, specifically including: calculating the inductance matrix of the winding-free motor, and performing matrix transformation to obtain a winding configuration matrix, wherein the winding configuration matrix includes the number of turns of each circuit in each slot.

[0017] In at least one possible implementation, the location tracking step includes:

[0018] The collected rotor position signal is filtered and denoised after decimal transformation, and then sent to the finite element coupled circuit combination model.

[0019] In at least one possible implementation, the model implementation step includes:

[0020] Creating a custom simulation block according to a preset resolution, the simulation block comprising: a finite element coupled circuit combination model, and a winding rotor induction motor connected to the finite element coupled circuit combination model, wherein the sub-end of the motor stator is connected to a controlled voltage source, the terminal of the rotor is connected to a computer whose external resistance can be adjusted in real time, and external connections are set at the input and output ends of the simulation block according to a set number of motor circuits, the external connections including an input voltage and an output current;

[0021] The input voltage state and the output current state corresponding to the motor circuit are expressed as a matrix representing a standard state space and stored in a preset lookup table;

[0022] In each set time step, the inductance matrix used to calculate the electromagnetic torque is obtained by interpolation method according to the rotor position signal and the lookup table.

[0023] In at least one possible implementation, the time step selection step includes: setting the time step in microseconds.

[0024] In at least one possible implementation, the monitoring and prediction results output after simulation are sent to the oscilloscope in waveform form from the analog output channel of the digital twin processing unit.

[0025] In at least one possible implementation, the simulation platform is specifically used to modify the execution state of the program, including running, stopping and pausing; to change the operating parameters of the finite element coupled circuit combination model of the winding rotor induction motor in real time; and to receive data after the finite element coupled circuit combination model of the winding rotor induction motor is run.

[0026] In at least one possible implementation method, the programming step includes: using a simulation platform to complete the programming of a finite element coupled circuit combination model of a winding rotor induction motor based on a real-time digital twin, and passing the programming results to a digital twin processing unit for compilation and execution.

[0027] The main design concept of the present invention is to construct a finite element coupled circuit combination mathematical model for motor state prediction without considering factors such as magnetic saturation, iron loss and capacitance effect in the motor model, and build a corresponding hardware system, and realize simulation through the aforementioned model and hardware system, and use it to monitor and predict the motor state. Specifically, the mathematical model takes into account the special geometric shape and winding structure of the motor, and can be flexibly adjusted through convenient matrix transformation, and any number of search coils and eccentricity and other faults can be added therein, so it is particularly suitable for real-time fault prediction of winding rotor induction motors. The present invention has strong detection adaptability, can greatly reduce economic losses and safety accidents caused by motor failures, and opens up a new way for real-time monitoring of motor status and fault prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described below with reference to the accompanying drawings, in which:

[0029] Figure 1 A flow chart of the overall idea of ​​the motor state monitoring and prediction method based on finite element and digital twin provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] Embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0031] The present invention proposes an embodiment of a motor state monitoring and prediction method based on finite element and digital twin. Specifically, Figure 1 shown, including:

[0032] Step S1, constructing a finite element coupled circuit combination model of a winding rotor induction motor, specifically including the following: a model construction link, a circuit identification link, and a finite element calculation link of an inductance matrix;

[0033] Step S2, building a hardware system for motor state monitoring and prediction based on finite element and digital twin, specifically including: connecting the digital twin processing unit with the simulation platform data for real-time simulation, and inputting the sampled signals of the motor voltage and current acquisition circuit into the digital twin processing unit, and detecting the rotor position signal through the angular displacement sensor and inputting it into the digital twin processing unit; wherein the digital twin processing unit is configured as a finite element coupled circuit combination model of the winding rotor induction motor, and is used to output the prediction result;

[0034] Step S3, using the simulation platform to perform real-time processing to monitor and predict the motor state, specifically including the following: programming link, position tracking link, model implementation link and time step selection link.

[0035] Furthermore, the model building step includes:

[0036] Obtaining a total magnetic flux received by any motor circuit from other motor circuits, specifically, the total magnetic flux is equal to the sum of the magnetic fluxes received from other motor circuits; and based on a given motor circuit, the magnetic flux received from another motor circuit is related to the current flowing in the transmitter and the magnetic coupling between the two circuits; wherein the magnetic coupling is a function of the rotor angular displacement;

[0037] A dynamic model is obtained by using a single matrix equation to cover the above relationship between magnetic flux and current, and combining it with the voltage generated by the change of magnetic flux of any motor circuit; wherein the matrix equation includes the pre-obtained inductance matrix function;

[0038] The electromagnetic torque is calculated using the dynamic model, and the back electromotive force is obtained based on a first matrix that predefines the coupling relationship between the search coil and other motor circuits.

[0039] Furthermore, the circuit identification step includes:

[0040] Three winding coils are arranged on the motor stator and the motor rotor respectively, and a search coil is arranged around the stator teeth, so that the seven coil circuits constitute the hardware architecture of the finite element coupled circuit combination model of the winding rotor induction motor.

[0041] Furthermore, the finite element calculation process of the inductance matrix includes:

[0042] The inductance matrix functions of multiple rotor positions are calculated based on two-dimensional finite elements, and a lookup table is established, specifically including: calculating the inductance matrix of the winding-free motor, and performing matrix transformation to obtain a winding configuration matrix, wherein the winding configuration matrix includes the number of turns of each circuit in each slot.

[0043] In at least one possible implementation, the location tracking step includes:

[0044] The collected rotor position signal is filtered and denoised after decimal transformation, and then sent to the finite element coupled circuit combination model.

[0045] Furthermore, the model implementation process includes:

[0046] Creating a custom simulation block according to a preset resolution, the simulation block comprising: a finite element coupled circuit combination model, and a winding rotor induction motor connected to the finite element coupled circuit combination model, wherein the sub-end of the motor stator is connected to a controlled voltage source, the terminal of the rotor is connected to a computer whose external resistance can be adjusted in real time, and external connections are set at the input and output ends of the simulation block according to a set number of motor circuits, the external connections including an input voltage and an output current;

[0047] The input voltage state and the output current state corresponding to the motor circuit are expressed as a matrix representing a standard state space and stored in a preset lookup table;

[0048] In each set time step, the inductance matrix used to calculate the electromagnetic torque is obtained by interpolation method according to the rotor position signal and the lookup table.

[0049] Furthermore, the time step selection step includes: setting the time step in microseconds.

[0050] Furthermore, the monitoring and prediction results output after simulation are sent to the oscilloscope in the form of waveforms through the analog output channel of the digital twin processing unit.

[0051] Furthermore, the simulation platform is specifically used to modify the execution state of the program, including running, stopping and pausing; to change the operating parameters of the finite element coupled circuit combination model of the winding rotor induction motor in real time; and to receive data after the finite element coupled circuit combination model of the winding rotor induction motor is run.

[0052] Furthermore, the programming process includes: using a simulation platform to complete the programming of a finite element coupled circuit combination model of a winding rotor induction motor based on real-time digital twins, and transmitting the programming results to a digital twin processing unit for compilation and execution.

[0053] In combination with the above embodiments, the following is a detailed description:

[0054] (I) Constructing a finite element coupled circuit combination model of a winding rotor induction motor

[0055] 1. Model building

[0056] Wound induction motors and other types of motors contain N circuits. The total flux in any circuit is the sum of the fluxes it receives from all other circuits. (Each quantity introduced is instantaneous, no reference transformation is used)

[0057]

[0058] The magnetic flux received on a given circuit from another circuit depends on the current i flowing in the transmitter and the magnetic coupling between the two, which is a function of the rotor angular displacement θ.

[0059] φ ij =L ij (θ)i j (2)

[0060] All of the above equations are contained in a single matrix equation:

[0061] [φ]=[L(θ)][i] (3)

[0062] Any change in magnetic flux in a circuit generates a voltage v:

[0063]

[0064] Combining equation (3) and equation (4), we get the dynamic system of equation (5).

[0065]

[0066] Electromagnetic torque T em The calculation method is:

[0067]

[0068] For a search coil or any open circuit, use equation (7) instead to calculate the back EMF v m , to improve the numerical stability of the ordinary differential equation solver. The matrix [L m ] defines the coupling between the search coil and other current carrying circuits.

[0069]

[0070] In summary, the identification of the model requires knowing the inductance matrix function [L(θ)] and then inverting and differentiating it. Using finite element software to identify [L(θ)] makes the CFE-CC model universal. It uses a lookup table to store the inverse and derivative inductance matrices for many discrete positions of the rotor.

[0071] The pre-calculated inductance matrix function can be used to take into account magnetic saturation by applying a correction factor to the calculated flux linkage. The performance and calculation time depend on the type of machine and the number of factors selected to calculate.

[0072] 2. Circuit identification

[0073] Specifically, three winding coils (a s ,b s ,c s ), three winding coils are arranged on the motor rotor (a r ,b r ,c r ). In order to measure the voltage passing through the stator during operation, a search coil w is arranged around the stator teeth. s , seven coil circuits constitute the finite element-coupled circuit model of the winding rotor induction motor.

[0074] The resistance matrix [R] of the winding rotor induction motor is a diagonal matrix. Each terminal is accessible, but since the search coil is open circuit, the resistance values ​​of the remaining coils can be easily measured except the search coil. Therefore, in this special case, equation (7) is used instead of equation (5).

[0075]

[0076] The inductance matrix [L], for any given rotor position, has the following form. The magnetic coupling of the search coils is also included in a separate matrix [L w ]middle.

[0077]

[0078] [L w ]=[L wsas L wsbs Lwscs L wsar L wsbr L wscr ] (10)

[0079] 3. Finite element calculation of inductance matrix

[0080] The calculation of [L(θ)] uses finite element software and the exact machine geometry and winding scheme. Since the FEM results are used as pre-computed lookup tables, the computational time required to solve the finite element domain is independent of the computational time of the CFE-CC model at each time step.

[0081] Since the geometry of the winding rotor induction motor is more suitable for 2D finite element software, in the present work, 2D finite element software is selected to calculate [L(θ)] more quickly. By changing the calculated inductance matrix, their respective effects are added to the model.

[0082] Since the number of pole pairs of the motor is 2 and the motor is symmetrical, the motor model can be simplified to half of the entire motor. All magnetic materials are assumed to be linear, so magnetic saturation is not considered. The meshing is done using the software's automatic meshing tool. The mesh size is determined by comparing the results of each meshing until the accuracy change is obvious. Once the finite element model preparation is completed, the procedure for identifying the magnetic coupling begins. The lookup table is established by calculating the inductance matrix [L] for multiple rotor positions.

[0083] First, calculate the inductance matrix [L(θ)] of the motor without windings, that is, the inductance matrix of the slot [L slot (θ)]. In this way, any winding configuration can be generated without restarting the finite element process, saving time, simply by applying the following matrix transformation:

[0084] [L(θ)]=[N] T [L slot (θ)][N] (11)

[0085] [N] is the winding configuration matrix. It contains the number of turns of each circuit in each slot. Equation (12) shows how to build the winding configuration matrix, where n c is the total number of circuits, n s is the total number of slots.

[0086]

[0087] In total, an average of 1440 positions are calculated over a 180° rotation, i.e., each stator slot contains 60 positions or an angular step of 0.125°. At this point, the physical memory required to store the lookup table needs to be considered. Since the wound rotor induction motor has 6 circuits without search coils, the total number of elements in the back-inductance matrix lookup table is 51840. If the torque calculation is included, another lookup table of the same size is required to calculate the coefficients.

[0088] (II) Hardware system construction

[0089] 1. Equipment selection

[0090] The digital twin processing unit is the entry-level DRTS of the OP4510 OPAL-RT. Real-time calculations are performed by a dedicated multi-core CPU with a time step of 2μs. For applications that require smaller time steps, a higher-end DRTS or FPGA-based platform is required.

[0091] OP4510 is connected to a computer with RT-LAB installed via a network data cable. RT-LAB is a software environment for real-time simulation that can be connected to the DRTS of OPAL-RT. RT-LAB can:

[0092] (1) Modify the execution status of the program (run, stop, pause);

[0093] (2) Changing the parameters of the running model in real time;

[0094] (3) Receive data from the running model.

[0095] 2. Measurement of voltage and current:

[0096] The measured voltage is used as the input of the model, and the measured current is used for model validation and comparison with the output of the RTDT (real-time digital twin) model. A printed circuit board is designed to directly collect current and voltage from the motor, and then the signal is sampled using the analog input channel of the DRTS.

[0097] 3. Angular displacement sensor

[0098] The angular position of the rotor (angular displacement of the rotor) needs to be retrieved from the lookup table of the inductance matrix, which is measured by an absolute shaft encoder. The measured position is encoded as a 12-bit signal and sent through the parallel communication interface. Each bit is acquired by the digital input channel of the DRTS and converted from binary to decimal in software.

[0099] 4. Hardware structure

[0100] The CFE-CC model is executed in real time as the actual machine runs, and all estimated waveforms can be displayed on an oscilloscope through the analog output channels of the DRTS.

[0101] (III) Implementation process of real-time simulator:

[0102] 1. Programming

[0103] In practice, RTDT programming can be done in the MATLAB / Simulink environment using RT-LAB, which generates C code from a Simulink diagram and sends it to DRTS for compilation and execution.

[0104] 2. Location tracking

[0105] The collected position signal is a noisy discrete signal. After decimal transformation, the position signal is filtered and denoised and then sent to the aforementioned model, which will obtain the velocity impulse that affects its prediction.

[0106] The closed-loop transfer function of the position signal filtering and noise reduction control loop is:

[0107]

[0108] where k p and k i is the proportional and integral gain of the position signal noise reduction PI controller. The measured position is converted from a sawtooth waveform to a ramp waveform in advance. The difference between the measured position and the estimated position is sent to the PI controller and the integrator. Therefore, the error in steady state is theoretically reduced to zero. The gain of the controller is preferably high enough to track the speed change of the machine.

[0109] 3. Real-time CFE-CC model implementation

[0110] System resolution is achieved through solvers provided by the software, which allow us to solve the power system equations with a hybrid approach of pure state space and nodal methods. It is built on the (simulated power systems) SimPowerSystems (SPS) package, so Simulink diagrams are designed in the same way by using SPS blocks (resistors, sources, inductors, etc.).

[0111] A custom block for SSN resolution is implemented in the CFE-CC model, which contains the CFE-CC model equations and the external components connected to it, the winding rotor induction motor, the sub-end of the stator is connected to the controlled voltage source, which is the stator voltage measured on the physical machine. The terminal of the rotor is connected to the computer where the external resistance can be adjusted in real time. By placing the motor in a separate node group instead of an external component, it has a decoupling effect from the SSN resolution, so even if the external resistance is large, strong stability can be achieved.

[0112] In order to implement such a model at SSN resolution, it is necessary to first select which circuits are available for external connections of the target of study, i.e., the terminals of the Simulink block (simulation block) of the motor. The winding rotor induction motor has six circuits, all of which can be used for external connections, so the vector v m and i out Each has six units. They are the input voltage and output current at both ends of the Simulink block. Because the search coil w s It is an open circuit, no current flows, so the search coil w s Not considered a circuit.

[0113] [φ]=[φ as φ bs φ cs φ ar φ br φ cr ] T (14)

[0114] [v m ]=[v as v bs v cs v ar v br v cr ] T (15)

[0115] [i out ]=[i as i bs i cs i ar i br i cr ] T (16)

[0116] The kinetic model of equation (5) is rearranged into the standard state space form. For simplicity, the following equation omits the θ dependence.

[0117] [φ]=[A][φ]+[B][v in ] (17)

[0118] [i out ]=[C][φ] (18)

[0119] [A]=-[R][L] -1 (19)

[0120] [B]=[I] (20)

[0121] [C]=[L] -1 (twenty one)

[0122] Matrix [A d (θ)],[B d (θ)] and [L(θ)] -1 Stored in a lookup table. In each time step, the exact matrix is ​​calculated by interpolating between two adjacent matrices in the lookup table through the angular movement measurement. Using the same type of storage and interpolation method, the derivative inductance matrix for torque calculation is obtained through equation (6).

[0123] Now, for the search coil, since there is no current, that is, no state, it is not included in the power system, and is calculated using formula (7).

[0124] 4. Time step selection

[0125] The execution time step of the program should be chosen as small as possible. If DRTS does not complete all calculations within the specified time step, an overflow occurs. In this case, a time step is skipped, resulting in an erroneous prediction. RT-LAB provides information about the computational requirements of the program. Based on this, a time step of 6 μs was selected. Small time steps are important, especially when the circuit of a CFE-CC model is connected to a large external resistor. Large external resistances create a stiff state-space system, so smaller time steps are required to keep the solver stable. The SSN solver improves stability compared to traditional methods, but the solution accuracy decreases when the time step is too large.

[0126] In summary, the main design concept of the present invention is that, without considering the factors such as magnetic saturation, iron loss and capacitance effect in the motor model, a finite element coupled circuit combination mathematical model for motor state prediction is constructed, and a corresponding hardware system is built, and simulation is realized through the aforementioned model and hardware system, and it is used to monitor and predict the motor state, specifically, the mathematical model takes into account the special geometric shape and winding structure of the motor, and can be flexibly adjusted through convenient matrix transformation, and any number of search coils and eccentricity and other faults can be added therein, so it is particularly suitable for real-time fault prediction of winding rotor induction motor. The present invention has strong detection adaptability, can greatly reduce the economic losses and safety accidents caused by motor failures, and opens up a new way for real-time monitoring and fault prediction of the motor state.

[0127] In the embodiments of the present invention, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. A and B may be singular or plural. The character " / " generally indicates that the associated objects before and after are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c may represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, c may be single or multiple.

[0128] The above describes in detail the structure, features and effects of the present invention based on the embodiments shown in the drawings, but the above is only a preferred embodiment of the present invention. It should be noted that the technical features involved in the above embodiments and their preferred methods can be reasonably combined and matched into a variety of equivalent schemes by those skilled in the art without departing from or changing the design ideas and technical effects of the present invention; therefore, the present invention is not limited to the scope of implementation shown in the drawings, and all changes made in accordance with the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which still do not exceed the spirit covered by the specification and drawings, should be within the protection scope of the present invention.

Claims

1. A motor state monitoring and prediction method based on finite element and digital twin, characterized in that: include: Step S1, constructing a finite element coupled circuit combination model of a winding rotor induction motor, specifically including the following: a model construction link, a circuit identification link, and a finite element calculation link of an inductance matrix; The model building process includes: obtaining the total magnetic flux received by any motor circuit from other motor circuits, specifically, the total magnetic flux is equal to the sum of the magnetic flux received from other motor circuits; and based on a given motor circuit, the magnetic flux received from another motor circuit is related to the current flowing in the transmitter and the magnetic coupling between the two circuits; wherein the magnetic coupling is a function of the rotor angular displacement; using a single matrix equation to cover the above magnetic flux and current relationship, and combining the voltage generated by the magnetic flux change of any motor circuit to obtain a dynamic model; wherein the matrix equation includes a pre-obtained inductance matrix function; using the dynamic model to calculate the electromagnetic torque, and obtaining the back electromotive force based on a first matrix that pre-defines the coupling relationship between the search coil and other motor circuits; The circuit identification link includes: arranging three winding coils on the motor stator and the motor rotor respectively, and arranging a search coil around the stator teeth, so that the seven coil circuits constitute the hardware architecture of the finite element coupled circuit combination model of the winding rotor induction motor; The finite element calculation link includes: calculating the inductance matrix function of multiple rotor positions based on two-dimensional finite elements, and establishing a lookup table, specifically including: calculating the inductance matrix of the winding-free motor, and performing matrix transformation to obtain a winding configuration matrix, wherein the winding configuration matrix includes the number of turns of each circuit in each slot; Step S2, building a hardware system for motor state monitoring and prediction based on finite element and digital twin, specifically including: connecting the digital twin processing unit with the simulation platform data for real-time simulation, and inputting the sampled signals of the motor voltage and current acquisition circuit into the digital twin processing unit, and detecting the rotor position signal through the angular displacement sensor and inputting it into the digital twin processing unit; wherein the digital twin processing unit is configured as a finite element coupled circuit combination model of the winding rotor induction motor, and is used to output the prediction result; Step S3, using the simulation platform to perform real-time processing to monitor and predict the motor state, specifically including the following: programming link, position tracking link, model implementation link and time step selection link; Among them, the model implementation link includes: creating a customized simulation block according to a preset resolution, the simulation block includes: a finite element coupled circuit combination model, and a winding rotor induction motor connected to the finite element coupled circuit combination model, wherein the sub-end of the motor stator is connected to a controlled voltage source, and the terminal of the rotor is connected to a computer whose external resistance can be adjusted in real time, and according to the set number of motor circuits, external connections are set at the input and output ends of the simulation block, and the external connections include input voltage and output current; the input voltage state and output current state corresponding to the motor circuit are expressed as a matrix representing the standard state space, and stored in a preset lookup table; in each set time step, according to the rotor position signal and the lookup table, the inductance matrix used to calculate the electromagnetic torque is obtained by interpolation.

2. The motor state monitoring and prediction method based on finite element and digital twin according to claim 1 is characterized in that: The location tracking process includes: The collected rotor position signal is filtered and denoised after decimal transformation, and then sent to the finite element coupled circuit combination model.

3. The motor state monitoring and prediction method based on finite element and digital twin according to claim 1 is characterized in that: The time step selection step includes: setting the time step in microseconds.

4. The motor state monitoring and prediction method based on finite element and digital twin according to any one of claims 1 to 3, characterized in that: The monitoring and prediction results output after simulation are sent to the oscilloscope in waveform form through the analog output channel of the digital twin processing unit.

5. The motor state monitoring and prediction method based on finite element and digital twin according to any one of claims 1 to 3, characterized in that: The simulation platform is specifically used to modify the execution state of the program, including running, stopping and pausing; to change the operating parameters of the finite element coupled circuit combination model of the winding rotor induction motor in real time; and to receive data after the finite element coupled circuit combination model of the winding rotor induction motor is run.

6. The motor state monitoring and prediction method based on finite element and digital twin according to any one of claims 1 to 3, characterized in that: The programming process includes: using a simulation platform to complete the programming of a finite element coupled circuit combination model of a winding rotor induction motor based on a real-time digital twin, and transmitting the programming results to a digital twin processing unit for compilation and execution.