Multi-motor cooperative control method and device based on digital twinning and storage medium

Abstract

The application relates to a multi-motor cooperative control method and device based on digital twinning and a storage medium, and is applied to the technical field of multi-motor cooperative control, and comprises the following steps: a digital twinning model of a multi-motor cooperative application scene is constructed, virtual motors and real motors are mapped to each other, an improved position coupling model is introduced, the proportion of virtual output and real output results is balanced through the value of a control coefficient, the position coupling output is taken as a comparison object, the output of each motor is compared, a synchronization error is obtained, each motor is compensated for the synchronization error, meanwhile, since the running environment of a workshop generally has the characteristics of complexity and harshness, the anti-interference capability of system synchronization running is improved, an interference observation model is introduced into the digital twinning model, the comprehensive error caused by internal and external factors is estimated, and the estimated error is fed back to the motor control system for compensation, so that error elimination is completed, composite control is realized, and the precision of motor cooperative control is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of motor cooperative control, and particularly relates to a multi-motor cooperative control method and device based on digital twinning and a storage medium. BACKGROUND

[0002] With the rapid development of modern science and technology and the continuous improvement of industrial production automation, single motor control cannot meet the demand, especially for the use of large mechanical equipment which needs multiple motors to work together, so the application occasions of multi-motor cooperative control are more and more, in the industrial field, such as textile, chemical fiber and other industries need multiple rollers to cooperate, and multi-motor cooperative control will directly affect the reliability of production and the quality of products, so it has very important practical significance to ensure the stability of multi-motor cooperative control system and improve the accuracy of multi-motor system control.

[0003] In multi-motor cooperative control, there are mainly two control methods, mechanical method and electric control method, wherein the mechanical method is to realize by using gear, rack transmission or chain transmission, but it has been difficult to be applied to current large transmission equipment due to the disadvantages of complex structure, small transmission range and distance, poor flexibility, etc. At present, electric control method is more used, although the electric control method overcomes the shortcomings of the mechanical method and has the advantage of good anti-interference, but there are still the following two problems: one is that the cooperative control effect is poor, although many researchers have carried out extensive and in-depth research on the cooperative control structure and control algorithm, but due to the different technical parameters of each motor, and the influence of factors such as load change, motor parameter drift, transmission mechanism gap uncertainty, how to keep multiple motors running in high precision is the primary problem. SUMMARY

[0004] Therefore, the purpose of the present application is to provide a multi-motor cooperative control method and device based on digital twinning, to solve the problem that the multi-motor cooperative precision is not high due to the different technical parameters of each motor, and the influence of factors such as load change, motor parameter drift, transmission mechanism gap uncertainty in the prior art.

[0005] According to a first aspect of an embodiment of the present application, a multi-motor cooperative control method based on digital twinning is provided, comprising:

[0006] A digital twin model of the multi-motor cooperative control scene is constructed based on the multi-motor cooperative control scene in the real scene, and the digital twin model comprises a virtual motor model, an interference observation model and a position coupling model;

[0007] The control parameters of each motor in the real scene are obtained and input into the corresponding virtual motor model, the fuzzy PID controller based on the particle swarm algorithm outputs the theoretical output values of each motor in the virtual motor model, and the theoretical output values and the actual output values of each motor are input into the position coupling model, the actual output values of each motor output by the fuzzy PID controller based on the particle swarm algorithm in the real scene are obtained, and the actual output values are input into the position coupling model;

[0008] The position coupling model balances the proportion of the theoretical output values and the actual output values through the value of the control coefficient according to the theoretical output values and the actual output values of each motor, and obtains the coupling output;

[0009] The actual parameter perturbation of each motor in the real scene and the disturbance value caused by the load change are obtained, and the disturbance value is input into the disturbance observation model, and the disturbance output is obtained by the mathematical theoretical model of the controlled motor and the low-pass filter;

[0010] The actual output values of each motor are compensated by the disturbance output and the coupling output, and the compensation output values of each motor are obtained, and the compensation output values are sent to the controller of each motor in the real scene to control the output values of each motor.

[0011] Preferably,

[0012] The virtual motor model comprises:

[0013] A geometric model for representing the appearance shape, size and structural relationship of the motor;

[0014] A physical model for representing the force change, mechanical coupling characteristics and temperature coupling characteristics of the motor;

[0015] A behavior model for representing the motion state and action relationship of the motor during operation;

[0016] A rule model for representing the constraint conditions and motion range of the motor;

[0017] The geometric model of the motor is proportionally modeled and map rendered by 3dmax software; the physical model of the motor is simulated and calculated by ANSYS software; the behavior model and the rule model of the motor are set in the Unity3D environment;

[0018] The virtual motor model realizes the running state of the motor in the real scene by accessing the motor running signal in the real scene.

[0019] Preferably,

[0020] The disturbance observation model obtains the disturbance output through the mathematical theoretical model of the controlled motor and the low-pass filter, comprising:

[0021] Obtain the observed interference and noise values ​​of each motor in real-world scenarios;

[0022] The variable transmission function is calculated by using the selected time constant of the low-pass filter, and the transfer function of the controlled motor is obtained by using the variable transmission function and a mathematical theoretical model of the controlled motor;

[0023] Obtaining the equivalent interference value of the controlled motor through the mathematical theoretical model, transfer function and observed interference value of the controlled motor;

[0024] The interference output is obtained through the interference value, the equivalent interference value, the noise value, the mathematical theoretical model of the controlled motor, the transfer function of the controlled motor and the low-pass filter.

[0025] Preferably, it also includes:

[0026] The real-time status data of the motor in the real scene is obtained and sent to the digital twin model. The digital twin model uses wavelet decomposition to obtain multiple groups of data. The signal of each group of data represents a scale feature.

[0027] The multi-scale features are input into the SSAE network pre-built in the digital twin model to obtain the contribution rate of different scale features to different fault classification accuracies. Then, based on the contribution rate of different scale features to different fault classification accuracies, the decomposed signal is reconstructed to eliminate external noise interference in the real-time status data of the motor to obtain a reconstructed signal. The reconstructed signal is input into the SSAE network again to obtain the contribution rate of the reconstructed signal to different fault classification accuracies, thereby realizing the diagnosis of motor faults.

[0028] Preferably, it also includes:

[0029] The historical operating data of the motor in real-world scenarios is obtained and input into the digital twin model. The digital twin model extracts weak attenuation features from the historical operating data based on variational mode decomposition and Laplace eigenmaps.

[0030] The digital twin model inputs the time series of weak attenuation characteristics into a pre-built multi-motor state prediction neural network model of weak attenuation characteristics. The multi-motor state prediction neural network model of weak attenuation characteristics outputs the motor attenuation index time series, which is used to predict the development trend of the motor's weak attenuation characteristics.

[0031] Preferably,

[0032] The multi-motor state prediction neural network model with weak attenuation characteristics uses the LSTM neural network as the basic model structure, and adds a convolutional neural network to improve the prediction accuracy of the model.

[0033] According to a second aspect of the embodiment of the present application, a multi-motor cooperative control device based on digital twinning is provided, and the device comprises:

[0034] A model building module is configured to build a digital twinning model of a multi-motor cooperative control scene based on a multi-motor cooperative control scene in a real scene, wherein the digital twinning model comprises a virtual motor model, a disturbance observation model and a position coupling model.

[0035] A motor output acquisition module is configured to acquire control parameters of each motor in the real scene and input the control parameters into the corresponding virtual motor model, output theoretical output values of each motor in the virtual motor model by a fuzzy PID controller based on a particle swarm algorithm, and input the theoretical output values into the position coupling model, acquire actual output values of each motor output by the fuzzy PID controller based on the particle swarm algorithm in the real scene, and input the actual output values into the position coupling model.

[0036] A coupling output acquisition module is configured to balance the proportion of the theoretical output values and the actual output values by the value of a control coefficient according to the position coupling model based on the theoretical output values and the actual output values of each motor, and obtain a coupling output.

[0037] A disturbance output acquisition module is configured to acquire actual parameter perturbations of each motor in the real scene and disturbance values caused by load changes, input the disturbance values into the disturbance observation model, and obtain disturbance outputs by a mathematical theoretical model of the controlled motor and a low-pass filter.

[0038] A compensation output acquisition module is configured to compensate the actual output values of each motor by the disturbance outputs and the coupling outputs, obtain compensation output values of each motor, send the compensation output values to controllers of each motor in the real scene, and control the output values of each motor.

[0039] According to a third aspect of the embodiment of the present application, a storage medium is provided, and the storage medium stores a computer program, wherein the computer program is executed by a host computer to implement each step of the method for redesigning a logistics equipment based on digital twinning.

[0040] The technical solution provided by the embodiment of the present application can have the following beneficial effects:

[0041] The application constructs a digital twin model of a multi-motor cooperative application scene, maps a virtual motor with a real motor by acquiring parameters of the real motor, introduces an improved position coupling model in the digital twin model, so that the synchronization error of each motor is only related to the position coupling output of the twin motor, the calculation is simple, and the deviation coupling property is also possessed, the proportion of the motor simulation result of the twin model and the actual system running result is balanced by controlling the value of the coefficient, the position coupling output is taken as the comparison object, and the output of each motor is compared to obtain the synchronization error, the synchronization error compensation is performed on each motor, the interference observation model is introduced in the digital twin model to improve the anti-interference ability of the system synchronization operation, the comprehensive error caused by internal and external factors is estimated, and the estimated error is fed back to the motor control system for compensation, so as to complete error elimination and realize compound control, thereby improving the precision of motor cooperative control.

[0042] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the application. BRIEF DESCRIPTION OF DRAWINGS

[0043] The drawings incorporated into the specification and forming part of the specification, show embodiments consistent with the application, and together with the specification, serve to explain the principles of the application.

[0044] Figure 1 is a flowchart of a multi-motor cooperative control method based on digital twin according to an exemplary embodiment;

[0045] Figure 2 is a schematic diagram of a deviation coupling control principle according to another exemplary embodiment;

[0046] Figure 3 is a schematic diagram of a deviation coupling control structure according to another exemplary embodiment;

[0047] Figure 4 is a schematic diagram of a disturbance observation model according to another exemplary embodiment;

[0048] Figure 5 is a system schematic diagram of a multi-motor cooperative control device based on digital twin according to another exemplary embodiment;

[0049] In the drawings: 1-model construction module, 2-motor output acquisition module, 3-coupling output acquisition module, 4-disturbance output acquisition module, 5-compensation output acquisition module. DETAILED DESCRIPTION

[0050] The exemplary embodiments will be described in detail herein with reference to the attached drawings. The description below refers to the accompanying drawings, which show, by way of example, specific embodiments with which the inventive concept can be practiced. The following detailed description is not intended to limit the scope of the invention, as claimed, but is intended to be illustrative thereof. Rather, the following detailed description is intended to describe devices and methods in accordance with some aspects of the invention as detailed in the appended claims.

[0051] Embodiment One

[0052] Figure 1 is a flowchart of a multi-motor cooperative control method based on digital twinning according to an exemplary embodiment, as shown in Figure 1 , the method comprises:

[0053] S1, constructing a digital twin model of a multi-motor cooperative control scene based on a multi-motor cooperative control scene in a real scene, the digital twin model comprising a virtual motor model, a disturbance observation model, and a position coupling model;

[0054] S2, obtaining control parameters of each motor in the real scene and inputting them into the corresponding virtual motor model, outputting theoretical output values of each motor in the virtual motor model by a fuzzy PID controller based on a particle swarm algorithm, and inputting them into the position coupling model, obtaining actual output values of each motor output by the fuzzy PID controller based on the particle swarm algorithm in the real scene, and inputting them into the position coupling model;

[0055] S3, the position coupling model balances the proportion of theoretical output values and actual output values by controlling the value of the coefficient according to the theoretical output values and actual output values of each motor, and obtains a coupling output;

[0056] S4, obtaining actual parameter perturbations of each motor in the real scene and disturbance values caused by load changes, inputting the disturbance values into the disturbance observation model, and obtaining disturbance outputs by the mathematical theoretical model of the controlled motor and a low-pass filter;

[0057] S5, compensating the actual output values of each motor by the disturbance outputs and the coupling outputs, obtaining compensation output values of each motor, sending the compensation output values to the controllers of each motor in the real scene, and controlling the output values of each motor;

[0058] It can be understood that by summarizing the advantages and disadvantages of various control structures, the deviation coupling adopts a compensation control strategy, which is better in terms of comprehensive effect from start-up characteristics, disturbance suppression ability, applicable scope, and engineering implementation convenience, etc., but is easily affected by the uncertainty of the load and unknown disturbances, etc. Therefore, the original result is compensated based on the simulation results of the simulation model on the basis of the deviation coupling control, so as to compensate for the error generated by the multi-motor synchronization. The deviation coupling control strategy improves the precision of multi-motor synchronization control and has excellent synchronization effect because the disturbance from the motor does not affect the master motor and other slave motors. However, the compensation value of each motor needs to be calculated based on the error between the motor and all other motors during the running process. If the system contains N motors, the synchronization compensator needs to calculate N(N-1) errors. The mathematical model of the system is relatively complex, the online calculation amount is large, and when the load changes, the real-time performance of the control strategy is poor, the start-up time is too long in the early stage of synchronization control, and the synchronization accuracy cannot be well guaranteed. Therefore, the control structure is improved on the basis of the deviation coupling control structure, as shown in the accompanying Figure 3 The familiar twin model is represented in the lower solid box in the figure, and the entity motor has the same configuration and attributes as λ i (i=n) represents the coupling coefficient, the controller adopts a fuzzy PID controller based on particle swarm algorithm (PSO-FC-PID), the digital twin model includes a virtual motor model, a disturbance observation model, and a position coupling model. In the improved deviation coupling structure, the synchronization error of each motor is only related to the position coupling output of the twin motor, which is simple to calculate and has the same properties as the deviation coupling. The specific process is as follows:

[0059] The present application constructs a digital twin model of multi-motor cooperative control scene based on the multi-motor cooperative control scene in the real scene. The digital twin model includes a virtual motor model, a disturbance observation model, and a position coupling model. The virtual motor model realizes the virtual-real mapping between the virtual motor and the real motor by obtaining the input and operating parameters of the real motor. The theoretical output values of each motor in the virtual motor model are output by the fuzzy PID controller based on particle swarm algorithm and input into the position coupling model. The actual output values of each motor based on the fuzzy PID controller based on particle swarm algorithm in the real scene are obtained and input into the position coupling model. Although the virtual motor model and the real motor use the same fuzzy PID controller based on particle swarm algorithm to output the output values, the real motor is affected by various external disturbances, motor parameter drift, transmission mechanism gap uncertainty, and other factors, while the virtual motor model outputs the output values under the theoretical condition of the motor. After obtaining the virtual-real two output values of the motor, the proportion of the simulation results of the twin motor and the actual system running results is balanced by the value of the control coefficient η (η1+η2=1), and the coupled output is obtained, which is expressed as follows.

[0060]

[0061] In the formula, n represents the number of motors in the system; S vir represents the theoretical output value of each motor in the virtual motor model; s ave represents the coupling output; S i represents the actual output value of each motor in the real scene;

[0062] It is worth emphasizing that the output values of the above-mentioned virtual motor model and the motors in the real scene are all output by the fuzzy PID controller based on the particle swarm algorithm, and the principle is shown in the attached Figure 2 To better and smoothly output the final speed of each motor, a fuzzy PID control algorithm based on particle swarm optimization is provided in the control strategy module to optimize the controller of each motor. The fuzzy rule base of the fuzzy PID controller of the multi-motor cooperative control system is optimized using the particle swarm algorithm, and the update speed and position of the particle swarm algorithm are dynamically adjusted, so that the controller has good adaptive ability.

[0063] After obtaining the coupling output, considering that the running environment of the workshop generally has complex and harsh characteristics, in order to improve the anti-interference ability of the synchronous operation of the system, an interference observation model is introduced in the digital twin model to realize compound control. During the simulation running of the twin motor, the disturbance value u caused by the actual parameter perturbation and the load change is collected by the data perception and collection system on the actual multi-motor running system, and the error is estimated through the interference observation model. The estimation result is taken as the interference output, the interference output is compensated to the control loop of each motor, thereby improving the stability of the system and improving the synchronization performance of the multi-motor lifting. The actual output value of each motor is compensated through the interference output and the coupling output, and the compensation output value of each motor is obtained. The compensation output value is sent to the controller of each motor in the real scene, and the output value of each motor is controlled.

[0064] The application constructs a digital twin model of a multi-motor collaborative application scene, maps a virtual motor with a real motor by acquiring parameters of the real motor, introduces an improved position coupling model in the digital twin model, so that the synchronization error of each motor is only related to the position coupling output of the twin motor, the calculation is simple, and the bias coupling property is also possessed, the proportion of the motor simulation result of the twin model and the actual system running result is balanced by controlling the value of the coefficient, the position coupling output is taken as a comparison object, compared with the output of each motor, the synchronization error is obtained, the synchronization error compensation is performed on each motor, the interference observation model is introduced in the digital twin model to improve the anti-interference ability of the system synchronization operation, the comprehensive error caused by internal and external factors is estimated, and the estimated error is fed back to the motor control system for compensation, so as to complete error elimination and realize composite control, thereby improving the precision of motor collaborative control.

[0065] Preferably,

[0066] The virtual motor model comprises:

[0067] A geometric model for representing the appearance shape, size and structural relationship of the motor;

[0068] A physical model for representing force variation, mechanical coupling characteristics and temperature coupling characteristics of the motor;

[0069] A behavior model for representing the motion state and action relationship of the motor in the running process;

[0070] A rule model for representing the constraint condition and motion range of the motor;

[0071] The geometric model of the motor is proportionally modeled and map rendering is restored by using 3dmax software; the physical model of the motor is simulated and calculated by using ANSYS software; the behavior model and the rule model of the motor are set in the Unity3D environment;

[0072] The virtual motor model realizes the running state of the motor in the real scene by accessing the motor running signal in the real scene;

[0073] It can be understood that various types of virtual motor models can be provided in the digital twin model for users to select, and the digital twin model of the motor is composed of a geometric model, a physical model, a behavior model and a rule model, wherein the geometric model represents the appearance shape, size and structural relationship of the motor; the physical model represents the force change, mechanical coupling characteristics and temperature coupling characteristics of the motor; the behavior model represents the motion state and action relationship of the motor in the running process; and the rule model represents the constraint conditions and motion range of the motor, etc. The geometric model of the motor is modeled and rendered in proportion by 3dmax software; the physical model of the motor is simulated and calculated by ANSYS software; and the behavior model and the rule model of the motor are set in the Unity3D environment. The virtual motor model is mapped with the physical motor, and the virtual motor model can reflect the running state of the physical motor in real time by accessing the running signal of the physical motor.

[0074] Preferably,

[0075] The disturbance observation model obtains the disturbance output through the mathematical theoretical model of the controlled motor and the low-pass filter, and the disturbance observation model comprises:

[0076] The observation disturbance value and the noise value of each motor in the real scene are obtained;

[0077] The variable transmission function is calculated through the time constant of the selected low-pass filter, and the transfer function of the controlled motor is obtained through the variable transmission function and the mathematical theoretical model of the controlled motor;

[0078] The equivalent disturbance value of the controlled motor is obtained through the mathematical theoretical model of the controlled motor, the transfer function and the observation disturbance value;

[0079] The disturbance output is obtained through the disturbance value, the equivalent disturbance value, the noise value, the mathematical theoretical model of the controlled motor, the transfer function of the controlled motor and the low-pass filter;

[0080] It can be understood that the disturbance observation model is as shown in the accompanying drawings, Figure 4 Wherein u represents input (the disturbance value u caused by actual parameter perturbation and load change, etc. is collected through a data perception collection system on an actual multi-motor running system), G(s) represents the transfer function of the controlled object (the controlled motor), G n (s) represents the mathematical theoretical model of the controlled object (the controlled motor), d represents the equivalent disturbance, ζ represents the measured noise, represents the observed disturbance, Q(s) represents the filter, and c represents the output, that is, e Figure 3 in the accompanying drawings, d That is, the disturbance output;

[0081] Therefore, the total input-output relationship of the above disturbance observation model is:

[0082] c(s) = G uc (s)u + G dc (s)d + G ξc (s)ξ

[0083] wherein:

[0084]

[0085]

[0086]

[0087] From the above two equations, the filter Q(s) in the interference observation model determines the dynamic performance of the entire interference observation model. The higher the order and the wider the bandwidth, the better the effect of suppressing external interference. However, too high an order will affect the sensitivity of the model and cause underdamping, leading to system instability. Therefore, selecting a reasonable low-pass filter order can filter out high-frequency noise while observing effective low-frequency interference.

[0088] Therefore, the filter Q(s) should be a low-pass filter, and the expression of the low-pass filter is:

[0089]

[0090] wherein, N represents the order of the low-pass filter, M represents the relative order, k represents the coefficient, and τ represents the time constant.

[0091] Since the mathematical theoretical model G n (s) of the controlled motor is known, and

[0092] G(s) = G n (s) [1 + Δ(s)]

[0093]

[0094] wherein, Δ(s) represents a high-frequency perturbation, which is a variable transfer function, T d represents the delay time.

[0095] Therefore, the transfer function G(s) of the controlled motor can be obtained.

[0096] Since the observed interference is known,

[0097]

[0098] Therefore, the equivalent interference d can be obtained, so according to:

[0099] c(s)=G uc (s)u+G dc (s)d+G ξc (s)ξ

[0100] The interference output e can be calculated d ;

[0101] After obtaining the coupled output s ave and interference output e d Afterwards, as attached Figure 3 As shown:

[0102]

[0103] Where s1 represents the actual system output value obtained by the PSO-FC-PID controller only without compensation by the position coupling model and the disturbance observation model; r represents the input of the system, which remains unchanged regardless of actual or simulation conditions.

[0104] After compensation by the position coupling model and the interference observation model, we get:

[0105]

[0106] Where s′1 represents the output value obtained by compensating the position coupling model and the interference observation model obtained in the twin system, that is, the output value expected by the final system.

[0107] Preferably, it also includes:

[0108] The real-time status data of the motor in the real scene is obtained and sent to the digital twin model. The digital twin model uses wavelet decomposition to obtain multiple groups of data. The signal of each group of data represents a scale feature.

[0109] The multi-scale features are input into the SSAE network pre-built in the digital twin model to obtain the contribution rate of different scale features to the accuracy of different fault classifications. The strength of the information implicit in each group of signals that affects the classification accuracy is determined. Then, based on the contribution rate of different scale features to the accuracy of different fault classifications, the wavelet coefficients corresponding to the signals with higher accuracy are amplified, while the wavelet coefficients corresponding to the signals with lower accuracy are attenuated. The decomposed signals are reconstructed to remove external noise interference from the real-time status data of the motor to obtain a reconstructed signal. The reconstructed signal is then input into the SSAE network again to obtain the contribution rate of the reconstructed signal to the accuracy of different fault classifications, thereby realizing the diagnosis of motor faults.

[0110] It can be understood that, due to the relatively harsh working environment of the motor, there is a possibility of failure. Although sensors are used in the feedback control of the system, if the feedback is not timely, once the motor fails, it will affect the cooperative control performance of the system, or even cause disastrous consequences to the system. Therefore, how to timely and effectively identify such failures and perform fault tolerance is of great significance to ensure product quality and ensure the safety and reliability of the system. In a multi-motor cooperative system, each motor is equipped with a sensor to sense the state information of the motor, and the motor state information is also uploaded to the digital twin model. Therefore, the digital twin model contains a large amount of historical operation data and real-time state data of each motor. The digital twin model obtains multi-scale features from real-time state data using wavelet decomposition. A string of data (signal) can obtain multiple groups of data (signals) after wavelet decomposition. Each group of signals represents a scale feature. The number of layers of the decomposed signals is determined according to the signal itself and the selected wavelet function. The multi-scale features are input into the SSAE network to obtain the contribution rate of different scales to the classification accuracy of different faults. The strength of the information affecting the classification accuracy in each group of signals is determined. According to the contribution of different scale signals to the classification accuracy of different faults, the wavelet coefficients corresponding to the signals with high accuracy are amplified, and the wavelet coefficients corresponding to the signals with low accuracy are attenuated. The decomposed signals are reconstructed to remove external noise interference of the real-time state data of the motor to obtain reconstructed signals. The reconstructed signals are input into the SSAE network to obtain the contribution rate of the reconstructed signals to the classification accuracy of different faults, and the motor fault is diagnosed.

[0111] Preferably, it further comprises:

[0112] The historical operation data of the motor in the real scene is obtained and input into the digital twin model. The digital twin model performs weak attenuation feature extraction based on variational modal decomposition and Laplace feature mapping on the historical operation data.

[0113] The digital twin model inputs the time series of the weak attenuation features into a pre-built multi-motor state prediction neural network model of the weak attenuation features. The multi-motor state prediction neural network model of the weak attenuation features outputs a time series of motor attenuation indicators, which is used to predict the development trend of the weak attenuation features of the motor.

[0114] It can be understood that when the fault prediction of the multi-motor is performed, the multi-motor historical operation data stored in the digital twin model is mainly applied, the multi-motor historical operation data is weakly attenuated feature extraction based on variational mode decomposition and Laplace feature mapping, so as to accurately extract the weakly attenuated feature of the motor, avoid the loss of key information and the interference of other irrelevant information, use the historical data memory and relationship analysis ability of the LSTM, and filter and retain the information of the attenuation process of the long-term data according to the characteristics of the motor weakly attenuated feature time sequence as the model input, the model output is the motor attenuation index time sequence, the development trend of the weakly attenuated feature of the motor is predicted, and the accurate prediction of the health state of the multi-motor is realized.

[0115] Preferably,

[0116] The weakly attenuated feature multi-motor state prediction neural network model takes the LSTM neural network as the basic structure of the model, and increases the convolutional neural network to improve the prediction accuracy of the model.

[0117] It can be understood that the LSTM neural network is taken as the basic structure of the model, and the convolutional neural network is increased to further improve the prediction accuracy of the model, and the weakly attenuated feature-based multi-motor state prediction model is established to predict the development trend of the weakly attenuated feature of the motor.

[0118] Embodiment two

[0119] The embodiment also discloses a system schematic diagram of a multi-motor cooperative control device based on digital twinning, as shown in the accompanying drawings. Figure 5 As shown, it comprises:

[0120] The model building module 1 is used to build a digital twin model of the multi-motor cooperative control scene based on the multi-motor cooperative control scene in the real scene, and the digital twin model comprises a virtual motor model, an interference observation model and a position coupling model.

[0121] The motor output acquisition module 2 is used to acquire the control parameters of each motor in the real scene and input them into the corresponding virtual motor model, output the theoretical output values of each motor in the virtual motor model based on the fuzzy PID controller of the particle swarm algorithm, and input them into the position coupling model, acquire the actual output values of each motor output by the fuzzy PID controller of the particle swarm algorithm in the real scene, and input them into the position coupling model.

[0122] The coupling output acquisition module 3 is used to balance the proportion of the theoretical output values and the actual output values by controlling the value of the coupling coefficient according to the theoretical output values and the actual output values of each motor according to the position coupling model, and obtain the coupling output.

[0123] The interference output acquisition module 4 is used for acquiring the interference value caused by the actual parameter perturbation of each motor in the real scene and the load change, inputting the interference value into an interference observation model, and obtaining the interference output through the mathematical theoretical model of the controlled motor and a low-pass filter.

[0124] The compensation output acquisition module 5 is used for compensating the actual output value of each motor through the interference output and the coupling output, obtaining the compensation output value of each motor, and sending the compensation output value to the controller of each motor in the real scene to control the output value of each motor.

[0125] It can be understood that the present application builds a digital twin model of a multi-motor cooperative control scene based on a multi-motor cooperative control scene in a real scene through the model building module 1, the digital twin model includes a virtual motor model, a disturbance observation model and a position coupling model; the motor output acquisition module 2 is used to acquire the control parameters of each motor in the real scene and input them into the corresponding virtual motor model, the fuzzy PID controller based on the particle swarm algorithm outputs the theoretical output values of each motor in the virtual motor model and inputs them into the position coupling model, the actual output values of each motor output by the fuzzy PID controller based on the particle swarm algorithm in the real scene are acquired and input into the position coupling model; the coupling output acquisition module 3 is used to acquire the theoretical output values and the actual output values of each motor according to the position coupling model, balance the proportion of the theoretical output values and the actual output values through the value of the control coefficient, and obtain the coupling output; the disturbance output acquisition module 4 is used to acquire the actual parameter perturbation of each motor in the real scene and the disturbance value caused by the load change, input the disturbance value into the disturbance observation model, and obtain the disturbance output through the mathematical theoretical model of the controlled motor and the low-pass filter; the compensation output acquisition module 5 is used to compensate the actual output values of each motor through the disturbance output and the coupling output, obtain the compensation output values of each motor, send the compensation output values to the controller of each motor in the real scene, and control the output values of each motor; the present application builds a digital twin model of a multi-motor cooperative application scene, acquires the parameters of the real motor, so that the virtual motor and the real motor are mapped to each other, in the digital twin model, an improved position coupling model is introduced, so that the synchronization error of each motor is only related to the position coupling output of the twin motor, the calculation is simple, and the same has the property of deviation coupling, the proportion of the simulation results of the twin motor and the actual system running results is balanced through the value of the control coefficient, the position coupling output is taken as the comparison object, compared with the output of each motor, the synchronization error is obtained, the synchronization error compensation is performed on each motor, at the same time, considering that the running environment of the workshop generally has the characteristics of complexity and harshness, in order to improve the anti-interference ability of the system synchronization running, a disturbance observation model is introduced in the digital twin model, the comprehensive error caused by internal and external factors is estimated, and the estimation error is fed back to the motor control system for compensation, so as to complete the error elimination and realize the compound control, thereby improving the precision of the motor cooperative control.

[0126] Embodiment three

[0127] The embodiment provides a storage medium, the storage medium stores a computer program, when the computer program is executed by a host computer, each step in the above method is realized;

[0128] It can be understood that the storage medium mentioned above can be a read-only memory, a disk or an optical disk.

[0129] It can be understood that the same or similar parts in the above-mentioned embodiments can be mutually referred to, and the content not described in detail in some embodiments can refer to the same or similar content in other embodiments.

[0130] It should be noted that in the description of the present application, the terms "first", "second" and the like are used only for descriptive purposes, and cannot be understood as indicating or implying relative importance. In addition, in the description of the present application, unless otherwise specified, the meaning of "a plurality of" is at least two.

[0131] Any process or method descriptions in flow charts or described herein otherwise can be understood as representing code modules, segments, or portions of code that include one or more executable instructions for performing specific logic functions or steps in the process, and that the various embodiments of the application include the additional implementation that the functions presented with no dependency on the storage and / or execution in the particular order discussed or illustrated. The processes, methods, and steps described in the figures and herein can be understood as generally being implemented in software, hardware, firmware, or a combination thereof.

[0132] It should be understood that the parts of the present application can be realized by hardware, software, firmware or their combination. In the above-mentioned embodiments, a plurality of steps or methods can be realized by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if realized by hardware, and as in another embodiment, it can be realized by any one or their combination of the following technologies known in the art: discrete logic circuit with logic gate circuit for implementing logic function on data signal, application specific integrated circuit with suitable combination logic gate circuit, programmable gate array (PGA), field programmable gate array (FPGA) and the like.

[0133] Those skilled in the art of the present technology can understand that all or part of the steps carried out by the above-mentioned embodiment method can be completed by a program instructing the relevant hardware, and the program can be stored in a computer readable storage medium, which includes one or a combination of the steps of the method embodiment when executed.

[0134] In addition, each functional unit in each embodiment of the present application can be integrated in one processing module, or each unit can be physically present separately, or two or more units can be integrated in one module. The above-mentioned integrated module can be realized in the form of hardware or in the form of software functional module. The integrated module, if realized in the form of software functional module and sold or used as an independent product, can also be stored in a computer readable storage medium.

[0135] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0136] In the description of the specification, reference to "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" means that a particular feature, structure, material, or characteristic being described is included in at least one embodiment or example of the application. The appearances of the above expressions in various places in the specification are not necessarily referring to the same embodiment or example. Moreover, the particular features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0137] Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are merely exemplary, and are not to be taken as limiting the present application, and that within the scope of the present application, changes, modifications, replacements, and variations of the above embodiments can be made by those skilled in the art.

Claims

1. A multi-motor cooperative control method based on digital twinning, characterized in that, The method comprises the following steps: A digital twin model of a multi-motor cooperative control scene is constructed based on a multi-motor cooperative control scene in a real scene, and the digital twin model comprises a virtual motor model, an interference observation model and a position coupling model; Control parameters of each motor in the real scene are obtained and input into the corresponding virtual motor model, a theoretical output value of each motor in the virtual motor model is output by a fuzzy PID controller based on a particle swarm algorithm, and the theoretical output value is input into the position coupling model, an actual output value of each motor output by the fuzzy PID controller based on the particle swarm algorithm in the real scene is obtained and input into the position coupling model; The position coupling model balances the proportion of the theoretical output value and the actual output value by the value of a control coefficient according to the theoretical output value and the actual output value of each motor, and obtains a coupling output; Actual parameter perturbations of each motor in the real scene and interference values caused by load changes are obtained, and the interference values are input into the interference observation model, and the interference observation model obtains an interference output by a mathematical theoretical model of the controlled motor and a low-pass filter; The actual output value of each motor is compensated by the interference output and the coupling output, and a compensation output value of each motor is obtained, and the compensation output value is sent to a controller of each motor in the real scene to control the output value of each motor.

2. The method of claim 1, wherein the virtual motor model comprises: a geometric model for representing the appearance shape, size and structural relationship of the motor; a physical model for representing force change, mechanical coupling characteristics and temperature coupling characteristics of the motor; a behavior model for representing the motion state and action relationship of the motor in the running process; a rule model for representing the constraint conditions and motion range of the motor; the geometric model of the motor is proportionally modeled and map rendered by 3dmax software; the physical model of the motor is simulated and calculated by ANSYS software; the behavior model and the rule model of the motor are set in the Unity3D environment; the virtual motor model realizes the running state of the motor in the real scene by accessing the motor running signal in the real scene.

3. The method of claim 1, wherein the interference observation model obtains the interference output by the mathematical theoretical model of the controlled motor and the low-pass filter comprises: obtaining observation interference values and noise values of each motor in the real scene; calculating a variable transmission function by the time constant of the selected low-pass filter, obtaining a transfer function of the controlled motor by the variable transmission function and the mathematical theoretical model of the controlled motor; obtaining an equivalent interference value of the controlled motor by the mathematical theoretical model of the controlled motor, the transfer function and the observation interference value; obtaining the interference output by the interference value, the equivalent interference value, the noise value, the mathematical theoretical model of the controlled motor, the transfer function of the controlled motor and the low-pass filter. Further comprising: real-time state data of the motor in the real scene is obtained and sent to the digital twin model, and the digital twin model obtains multiple groups of data by wavelet decomposition of the real-time state data of the motor, and each group of data represents a scale feature.

4. The method of claim 1, wherein, ​ ​ The multi-scale features are input into the pre-built SSAE network in the digital twin model to obtain the contribution rate of different scale features to the classification accuracy of different faults, and then the decomposed signals are reconstructed to eliminate the external noise interference of the real-time state data of the motor to obtain reconstructed signals.

5. The method of claim 1, wherein, Also includes: Obtain the historical operation data of the motor in the real scene and input it into the digital twin model, and the digital twin model extracts weak attenuation features based on variational modal decomposition and Laplace feature mapping; The digital twin model inputs the time series of weak attenuation features into the pre-built multi-motor state prediction neural network model of weak attenuation features, and the multi-motor state prediction neural network model of weak attenuation features outputs a time series of motor attenuation indicators, which is used to predict the development trend of the weak attenuation features of the motor.

6. The method of claim 5, wherein The multi-motor state prediction neural network model of weak attenuation features uses an LSTM neural network as the basic model structure and adds a convolutional neural network to improve the prediction accuracy of the model.

7. A multi-motor cooperative control device based on digital twinning, characterized by, The device includes: A model building module is configured to build a digital twin model of a multi-motor cooperative control scene based on a multi-motor cooperative control scene in a real scene, wherein the digital twin model includes a virtual motor model, an interference observation model, and a position coupling model. An output acquisition module is configured to acquire control parameters of each motor in a real scene and input them into the corresponding virtual motor model, output theoretical output values of each motor in the virtual motor model based on a fuzzy PID controller using a particle swarm algorithm, and input them into the position coupling model. A coupling output acquisition module is configured to balance the proportion of theoretical output values and actual output values by controlling the value of the coupling coefficient based on the position coupling model according to the theoretical output values and actual output values of each motor. An interference output acquisition module is configured to acquire actual parameter perturbations of each motor in a real scene and disturbance values caused by load changes, input the disturbance values into the interference observation model, and obtain interference outputs from the interference observation model through a mathematical theoretical model of the controlled motor and a low-pass filter. A compensation output acquisition module is configured to compensate the actual output values of each motor with the interference outputs and coupling outputs to obtain compensation output values of each motor, and send the compensation output values to the controllers of each motor in the real scene to control the output values of each motor.

8. A storage medium, characterized by The storage medium stores a computer program, and the computer program is executed by the host controller to implement each step in the multi-motor cooperative control method based on the digital twin as claimed in any one of claims 1-6.

Images