Online parameter identification method for permanent magnet synchronous motor

Abstract

The invention discloses a multi-parameter identification method for a permanent magnet synchronous motor. The method comprises the following steps: constructing a full-rank equation set based on a mathematical model of a PMSM in a d-q axis coordinate system; comparing a parameter output value of the motor in actual operation with a theoretical output value, and establishing a fitness function; parameter setting is carried out by adopting a Logistic chaotic mapping uniform initialization population strategy, and parameter optimization is carried out by utilizing an improved red-mouth blue-magpie optimization algorithm; and correcting the error of the fitness function until the maximum number of iterations is reached, and finally outputting an optimal identification parameter. Simulation results show that the improved Red-beak-Blue-Tapie optimization algorithm can identify the parameters of the permanent magnet synchronous motor more accurately and quickly, and the results are very close to actual motor parameter values.

Description

Technical Field

[0001] The technical field of a permanent magnet synchronous motor parameter identification method based on an improved red-billed blue magpie optimization algorithm.

[0002] The invention belongs to the field of permanent magnet synchronous motor control and relates to a permanent magnet synchronous motor parameter identification method. Background Art

[0003] Permanent magnet synchronous motor (PMSM) has been widely used in industrial drives, electric vehicles, aerospace and other fields due to its excellent high efficiency, high power density, low maintenance requirements and superior control performance. In order to achieve these advantages, advanced vector control and sensorless control technologies are often used. However, accurate motor parameters must be relied on to achieve precise motor control. In practical applications, the parameters of PMSM are affected by many factors, such as changes in the operating environment, different motor loads, and nonlinear effects of the magnetic field, which lead to dynamic changes in motor parameters. Therefore, accurate parameter identification technology has important theoretical and engineering significance.

[0004] Traditional parameter identification methods include recursive least squares (RLS), model reference adaptive system (MRAS), extended Kalman filter (EKF), etc. Although they perform well in some scenarios, they are challenging in dealing with complex motor models and nonlinear characteristics, and usually require a long identification time, which is not suitable for real-time application requirements. In recent years, artificial intelligence-based optimization algorithms such as particle swarm optimization (PSO) and genetic algorithm (GA) have attracted much attention due to their excellent global search capabilities and ability to handle high-dimensional problems. However, when applied to PMSM parameter identification problems, they usually face problems such as slow convergence speed and easy to fall into local optimal solutions.

[0005] Therefore, there is an urgent need to innovatively design a multi-parameter identification algorithm that can significantly improve the identification accuracy and real-time performance of permanent magnet synchronous motors to meet the challenges brought by complex working environments and dynamic load changes. Summary of the invention

[0006] In order to solve the above problems, the present invention provides a method for multi-parameter identification of permanent magnet synchronous motor, which includes the following key steps: based on the mathematical model of permanent magnet synchronous motor in dq axis coordinate system, construct a full rank equation group, which reflects the relationship between multiple parameters to be identified and synchronous motor parameters; compare the motor parameter output value during actual operation with the theoretical output value of the parameter, and establish a fitness function; combine the Logistic chaos mapping uniform initialization population strategy to set parameters; use the improved red-billed blue magpie optimization algorithm to optimize the position. Finally, error correction is performed according to the fitness function, and the iteration is stopped when the maximum number of iterations is reached, and the optimal identification parameters are output.

[0007] (1) First, obtain the state equations of the permanent magnet synchronous motor in the dq coordinate system. Discretize these equations and calculate the state equations when the d-axis current is zero (i d =0) and not zero (i d ≠0), thereby constructing the motor parameter identification model.

[0008] In the specific implementation process, the state parameters of the permanent magnet synchronous motor in the dq coordinate system are first collected, and then the state equation is established based on these data. The parameters to be identified include the stator resistance (R s ), d-axis inductance (L d ), q-axis inductance (L q ) and permanent magnet flux (v f ).

[0009] The above state equation is:

[0010]

[0011] In the formula, i d and i q are the d-axis current component and the q-axis current component, u d and i q are the d-axis voltage component and the q-axis voltage component, L d and L q is the d-axis inductance and q-axis inductance, R s is the stator resistance, ψ f is the permanent magnetic flux.

[0012] When considering the vector control strategy of a three-phase permanent magnet synchronous motor (PMSM), it is assumed that the rate of change of the d-axis and q-axis currents of the motor is approximately zero during stable operation, that is, di d / dt=0、di q / dt=0, and i d = 0 control strategy. This allows the PMSM discrete voltage equation in the dq coordinate system to be expressed as:

[0013]

[0014] When performing motor parameter identification, a current disturbance signal is applied to the motor stator d-axis in a short time to establish the motor's full-rank state equation group to achieve multi-parameter identification. d =0 and i d ≠0, collect the same data respectively, and obtain the full-rank state equation group. This method obtains the state equation group of the motor as follows:

[0015]

[0016] Here, u d0 (k) and u q0 (k),ω e (k), i q0 (k) represents d = kth sampling data under 0 control strategy; u d1 (k),u q1 (k), i d1 (k), i q1 (k) is the disturbance current i d1 The kth sampling data at time .

[0017] In the above formula, u d0 (k),u q0 (k),u d1 (k),u q1 (k) represents the actual output value of the motor, which is obtained by measuring the voltage sensor. Using the above fourth-order full-rank parameter identification model, the four parameters to be identified can be obtained, including the stator resistance (R s ), d-axis inductance (L d ), q-axis inductance (L q ) and permanent magnet flux (ψ f ).

[0018] (2) Then, the motor’s d-axis voltage, d-axis current, q-axis voltage, q-axis current, and motor speed are sampled N times. The motor identification model is used to calculate the theoretical output values ​​of the d-axis and q-axis voltages based on the motor state and input parameters at each sampling point. The fitness function is constructed by comparing the actual output values ​​with the theoretical values.

[0019] Construct the following fitness function:

[0020]

[0021] in, for The fitness function value of the corresponding individual, the actual output value y = [u d0 ,u q0 ,u d1 ,u q1 ] contains the d-axis and q-axis voltages measured under the two control strategies, where u d0 ,u q0 is in d =0 The d-axis and q-axis voltage values ​​measured under the control strategy, u d1 ,u q1 is in d ≠0 The measured d-axis and q-axis voltage values ​​under the control strategy; Theoretical output value is calculated based on a full-rank system of equations, where is ind =0 Theoretical values ​​of d-axis and q-axis voltages calculated under the control strategy, is in d ≠0 Theoretical values ​​of d-axis and q-axis voltages calculated under the control strategy.

[0022] (3) Then, the initial population is generated based on the Logistic map, and the current number of iterations t is set to 0; the initial position of each individual in the red-billed blue magpie population is calculated using the following formula:

[0023] x t+1 =rx t (1-x t )

[0024] X i =lb j +x t+1 ·(ub j -lb j ),i=1,2,…,N,

[0025] Where: x t is the chaotic vector of the tth iteration, x t+1 is the chaotic vector of the t+1th iteration, r is the mapping parameter; X i is the position of the i-th red-billed blue magpie in the search space, N is the number of populations, ub j and lb j They are the upper and lower limits of the search space, that is, the value range of the parameters to be identified.

[0026] (4) Get the initial position of the red-billed blue magpie and make corresponding corrections to the estimated values ​​of the parameters to be identified. The improved red-billed blue magpie optimization algorithm is used for motor parameter identification, which mainly includes three stages: finding food, attacking prey, and preserving food.

[0027] The food search phase: The mathematical model of food search is divided into two cases: small groups searching for food and large groups searching for food. The following is the corresponding mathematical formula:

[0028]

[0029] Where t represents the current iteration number, X i (t+1) represents the position of the i-th new search agent, p represents the number of hunts in a small group of red-billed blue magpies, q represents the number of hunts in a large group, X m represents the randomly selected mth individual, X rs (t) represents the randomly selected search agent in the current iteration, and Rand represents the variable used to generate random numbers.

[0030] At this stage, the selection numbers p and q were improved. They were originally generated by uniform random distribution. This method may limit the range of variation of the selection number in each iteration and cannot adapt well to the scenario of motor parameter identification. A method of randomly generating the selection numbers p and q using normal distribution was introduced, which is suitable for application scenarios such as motor parameter identification that require precise control of the selection number. The randomness and distribution of the selection number can promote the exploration of diversity of individuals in the solution space, help to find better parameter identification solutions, and improve the convergence speed and stability of the algorithm.

[0031] Attacking prey phase: In the attack phase, a tournament selection mechanism is introduced. The algorithm selects the best individual from the entire population through the tournament selection mechanism, and then decides whether to adopt a small group attack strategy or a group attack strategy to update the position of each individual based on the set probability Epsilon. Specifically, each individual first participates in a tournament, where individuals are randomly selected and compared based on fitness, and the winner will be used for subsequent updates. This process not only increases the diversity of the population, but also helps the algorithm to jump out of the local optimum and improve the global search capability. The formula for attacking prey can be summarized as:

[0032]

[0033] Among them: Epsilon is the probability of determining the attack strategy, X i is the current individual position, X food is the prey position, CF is the adjustment factor, where X p,mean and X q,mean is the average position recalculated based on the winner's position, X p,mean is the average of p individual positions randomly selected from a normal distribution in a small population, p is a random number between 2 and 5, X q,mean is the average value of q individual positions randomly selected from a normal distribution in a large population, q is a random number between 10 and N, and rand is a random vector.

[0034] Food preservation phase: In the Red-billed Blue Magpie Optimization (RBMO) algorithm, the formula of the food preservation phase is used to simulate the behavior of the Red-billed Blue Magpie storing food for future use. The purpose of this phase is to retain the information of the solution to help individuals find the global optimal solution during the optimization process. The following is the mathematical model of the food preservation phase:

[0035]

[0036] Where: X i (t+1) represents the position of individual i at time t+1; fitness old is the fitness value of the individual before updating its position; newis the fitness value of the individual after updating the position; X food (t) is the location of the food at time t, which is the optimal solution currently found.

[0037] For each individual, the individual's position is updated using different update strategies in the above formula. Finally, the fitness of the updated position is calculated and compared with the original fitness. If the new fitness is better, the individual's position and fitness are updated.

[0038] (5) The error between the actual output value and the theoretical output value is corrected according to the fitness function until the maximum number of iterations is reached and the iteration is stopped, and the optimal identification parameters are output.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The present invention improves the traditional red-billed blue magpie optimization algorithm, and uses the improved red-billed blue magpie optimization algorithm to realize multi-parameter identification of a permanent magnet synchronous motor. Among them, a large number of uniform initial populations are generated by using Logistic mapping, which can increase the diversity of the population. In order to improve the accuracy and convergence speed of identification, a method of randomly generating selection quantities p and q using normal distribution is introduced in each iterative food search stage, which effectively enhances the exploration ability of the algorithm, simulates the behavior of the red-billed blue magpie to flexibly adjust the group size in nature, increases the diversity of the search process, thereby avoiding the search from falling into the local optimal solution too early due to the convergence speed being too fast, and enhances the flexibility and stability of the algorithm to adapt to different optimization scenarios. Introducing a tournament selection mechanism in the attack prey stage can significantly improve the performance and efficiency of the algorithm. This method increases the propagation speed and probability of excellent solutions by selecting individuals with higher fitness to participate in reproduction, thereby accelerating the convergence of the algorithm. At the same time, it also helps to maintain population diversity and reduce the risk of the algorithm falling into the local optimal. In addition, the randomness of the tournament selection provides the algorithm with a stronger exploration ability, enabling it to jump out of the current search area and discover new and better solutions. The simulation results show that the method of the present invention can not only accurately identify the motor parameters, but also has the characteristics of fast convergence speed, strong stability and high efficiency, which makes the algorithm more effective and reliable when dealing with complex parameter identification problems, and has high practical value and promotion potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a control system block diagram of the permanent magnet synchronous motor of the present application.

[0042] Figure 2 This is a schematic diagram of a permanent magnet synchronous motor parameter identification method based on an improved red-billed blue magpie optimization algorithm in this application.

[0043] Figure 3The present invention is an example of using the permanent magnet synchronous motor parameter identification method to identify the change curve of the stator resistance.

[0044] Figure 4 The present invention is an example of using the permanent magnet synchronous motor parameter identification method to identify the change curve of the d-axis and q-axis inductances.

[0045] Figure 5 The present invention is an example of using the permanent magnet synchronous motor parameter identification method to identify the change curve of the permanent magnet flux. DETAILED DESCRIPTION

[0046] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0047] The present invention provides a method for multi-parameter identification of a permanent magnet synchronous motor, comprising the following key steps: constructing a full-rank equation group based on a mathematical model of the permanent magnet synchronous motor in a dq axis coordinate system; comparing the motor parameter output value during actual operation with the theoretical output value of the parameter, and establishing a fitness function; setting parameters in combination with a uniform initialization population strategy of a Logistic chaos map; and optimizing the position using an improved red-billed blue magpie optimization algorithm. Finally, error correction is performed according to the fitness function until the maximum number of iterations is reached and the iteration is stopped, and the optimal identification parameters are output.

[0048] (1) First, obtain the state equations of the permanent magnet synchronous motor in the dq coordinate system. Discretize these equations and calculate the state equations when the d-axis current is zero (i d =0) and not zero (i d ≠0), thereby constructing the motor parameter identification model.

[0049] In the specific implementation process, the state parameters of the permanent magnet synchronous motor in the dq coordinate system are first collected, and then the state equation is established based on these data. The parameters to be identified include the stator resistance (R s ), d-axis inductance (L d ), q-axis inductance (L q ) and permanent magnet flux (v f ).

[0050] The above state equation is:

[0051]

[0052] In the formula, i d and i q are the d-axis current component and the q-axis current component, u d and i q are the d-axis voltage component and the q-axis voltage component, L d and L qis the d-axis inductance and q-axis inductance, R s is the stator resistance, ψ f is the permanent magnetic flux.

[0053] According to Figure 1 When considering the vector control strategy of a three-phase permanent magnet synchronous motor (PMSM), it is assumed that the rate of change of the d-axis and q-axis currents of the motor is approximately zero during stable operation, that is, di d / dt=0、di q / dt=0, and i d = 0 control strategy. This allows the PMSM discrete voltage equation in the dq coordinate system to be expressed as:

[0054]

[0055] When performing motor parameter identification, a current disturbance signal is applied to the motor stator d-axis in a short time to establish the motor's full-rank state equation group to achieve multi-parameter identification. d =0 and i d ≠0, collect the same data respectively, and obtain the full-rank state equation group. This method obtains the state equation group of the motor as follows:

[0056]

[0057] Here, u d0 (k),u q0 (k),ω e (k), i q0 (k) represents d = kth sampling data under 0 control strategy; u d1 (k),u q1 (k), i d1 (k), i q1 (k) is the disturbance current i d1 The kth sampling data at time .

[0058] In the above formula, u d0 (k),u q0 (k),u d1 (k),u q1 (k) represents the actual output value of the motor, which is obtained by measuring the voltage sensor. Using the above fourth-order full-rank parameter identification model, the four parameters to be identified can be obtained, including the stator resistance (R s ), d-axis inductance (L d ), q-axis inductance (L q ) and permanent magnet flux (ψ f ).

[0059] (2) Then, the motor’s d-axis voltage, d-axis current, q-axis voltage, q-axis current, and motor speed are sampled N times. The motor identification model is used to calculate the theoretical output values ​​of the d-axis and q-axis voltages based on the motor state and input parameters at each sampling point. The fitness function is constructed by comparing the actual output values ​​with the theoretical values.

[0060] Construct the following fitness function:

[0061]

[0062] (3) The motor fitness function is optimized by the improved red-billed blue magpie optimization algorithm until the requirements are met, thereby obtaining the optimal parameter to be identified with the minimum fitness value.

[0063] like Figure 2 As shown in the figure, the steps of the method for realizing multi-parameter identification of permanent magnet synchronous motor based on the improved red-billed blue magpie optimization algorithm are as follows:

[0064] Step 1: Set the algorithm parameters in Matlab software, including the population size N (number of individuals), the dimension Dim of the search space (the number of parameters to be identified, in this invention, the parameters to be identified are (R s ,L d ,L q ,ψ f ), Dim=4), the upper limit ub and lower limit lb of the search space (the value range of the parameters to be identified).

[0065] Step 2: Generate the initial population based on the Logistic map, set the current iteration number t to 0, and calculate the initial position of each individual in the red-billed blue magpie population using the following formula:

[0066] x t+1 =rx t (1-x t )

[0067] X i =lb j +x t+1 ·(ub j -lb j ),i=1,2,…,N,

[0068] Where: x t is the chaotic vector of the tth iteration, x t+1 is the chaotic vector of the t+1th iteration, r is the mapping parameter; X i is the position of the i-th red-billed blue magpie in the search space, N is the number of populations, ub j and lb j They are the upper and lower limits of the search space, that is, the value range of the parameters to be identified.

[0069] Step 3, calculating the fitness function value of the individual according to the fitness function;

[0070] Step 4, using the improved red-billed blue magpie optimization algorithm to perform parameter identification on the permanent magnet synchronous motor;

[0071] Based on the initial position of the red-billed blue magpie, the estimated values ​​of the parameters to be identified are corrected accordingly. The motor parameter identification using the improved red-billed blue magpie optimization algorithm mainly includes three stages: finding food, attacking prey and preserving food.

[0072] The food search phase: The mathematical model of food search is divided into two cases: small groups searching for food and large groups searching for food. The following is the corresponding mathematical formula:

[0073]

[0074] Where t represents the current iteration number, X i (t+1) represents the position of the i-th new search agent, p represents the number of hunts in a small group of red-billed blue magpies, q represents the number of hunts in a large group, X m represents the randomly selected mth individual, X rs (t) represents the randomly selected search agent in the current iteration, and Rand represents the variable used to generate random numbers.

[0075] At this stage, the selection numbers p and q were improved. They were originally generated by uniform random distribution. This method may limit the range of variation of the selection number in each iteration and cannot adapt well to the scenario of motor parameter identification. A method of randomly generating the selection numbers p and q using normal distribution was introduced, which is suitable for application scenarios such as motor parameter identification that require precise control of the selection number. The randomness and distribution of the selection number can promote the exploration of diversity of individuals in the solution space, help to find better parameter identification solutions, and improve the convergence speed and stability of the algorithm.

[0076] Attacking prey phase: In the attack phase, a tournament selection mechanism is introduced. The algorithm selects the best individual from the entire population through the tournament selection mechanism, and then decides whether to adopt a small group attack strategy or a group attack strategy to update the position of each individual based on the set probability Epsilon. Specifically, each individual first participates in a tournament, where individuals are randomly selected and compared based on fitness, and the winner will be used for subsequent updates. This process not only increases the diversity of the population, but also helps the algorithm to jump out of the local optimum and improve the global search capability. The formula for attacking prey can be summarized as:

[0077]

[0078] Among them: Epsilon is the probability of determining the attack strategy, X i is the current individual position, X food is the prey position, CF is the adjustment factor, where X p,mean and X q,mean is the average position recalculated based on the winner's position, X p,mean is the average of the positions of p individuals randomly selected from a small population, p is a random number between 2 and 5, X q,mean is the average of the positions of q randomly selected individuals in the population, q is a random number between 10 and N, and rand is a random vector.

[0079] Food preservation phase: In the Red-billed Blue Magpie Optimization (RBMO) algorithm, the formula of the food preservation phase is used to simulate the behavior of the Red-billed Blue Magpie storing food for future use. The purpose of this phase is to retain the information of the solution to help individuals find the global optimal solution during the optimization process. The following is the mathematical model of the food preservation phase:

[0080]

[0081] Where: X i (t+1) represents the position of individual i at time t+1; fitness old is the fitness value of the individual before updating its position; new is the fitness value of the individual after updating the position; X food (t) is the location of the food at time t, which is the optimal solution currently found.

[0082] Step 5: Termination condition judgment

[0083] The error between the actual output value and the theoretical output value is corrected according to the fitness function until the maximum number of iterations is reached and the iteration is stopped, and the optimal identification parameters are output, that is, the individual with the minimum fitness value in the search process is selected as the optimal solution output.

[0084] In one example, the identification change curves of four parameters to be identified are as follows: Figure 3-Figure 5 As shown, the final identified values ​​of the four parameters to be identified and the actual values ​​and their errors are shown in the following table:

[0085] Motor parameters Identification value Actual value error Stator resistance / Ω 2.8755 2.8750 0.0571％ d, q axis inductance / H 0.008502 0.0085 0.0235％ Permanent magnet flux / Wb 0.175003 0.175 0.0017％

[0086] In this example, the four parameters to be identified are very close to the true values ​​of the motor parameters after convergence, which proves that the method of the present application has the ability to accurately identify the motor parameters.

[0087] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. The present invention provides a method for multi-parameter identification of a permanent magnet synchronous motor, comprising the following steps: Based on the mathematical model of the permanent magnet synchronous motor in the dq axis coordinate system, a full-rank equation group is constructed; the motor parameter output values ​​during actual operation are compared with the theoretical output values ​​of the parameters, and a fitness function is established; Combined with the Logistic chaotic mapping uniform initialization population strategy, parameter setting is performed; the improved red-billed blue magpie optimization algorithm is used for parameter optimization. Finally, error correction is performed according to the fitness function, and the iteration is stopped when the maximum number of iterations is reached, and the optimal identification parameters are output.

2. The method for identifying parameters of a permanent magnet synchronous motor according to claim 1, characterized in that: Obtain the state equations of the permanent magnet synchronous motor in the dq coordinate system. Discretize these equations and calculate the state equations when the d-axis current is zero (i d =0) and not zero (i d ≠0), thereby constructing the motor parameter identification model. Based on the vector control strategy of permanent magnet synchronous motor (PMSM), there are two control states, namely i d =0 and i d ≠0, collect the same data respectively, and obtain the full-rank state equation group. This method obtains the state equation group of the motor as follows: Here, u d0 (k) and u q0 (k),ω e (k), i q0 (k) represents d = kth sampling data under 0 control strategy; u d1 (k) and u q1 (k), i d1 (k), i q1 (k) is the disturbance current i d1 The kth sampling data at time .

3. The method for identifying parameters of a permanent magnet synchronous motor according to claim 2, characterized in that: The motor's d-axis voltage, d-axis current, q-axis voltage, q-axis current, and motor speed are sampled N times. The motor identification model is used to calculate the theoretical output values ​​of the d-axis and q-axis voltages based on the motor state and input parameters at each sampling point. The fitness function is constructed by comparing the actual output values ​​with the theoretical values. Construct the following fitness function: in, for The fitness function value of the corresponding individual, the actual output value y = [u d0 ,u q0 ,u d1 ,u q1 ] contains the d-axis and q-axis voltages measured under the two control strategies, where u d0 ,u q0 is in d =0 The d-axis and q-axis voltage values ​​measured under the control strategy, u d1 ,u q1 is in d ≠0 The measured d-axis and q-axis voltage values ​​under the control strategy; Theoretical output value is calculated based on a full-rank system of equations, where is in d =0 Theoretical values ​​of d-axis and q-axis voltages calculated under the control strategy, is in d ≠0 Theoretical values ​​of d-axis and q-axis voltages calculated under the control strategy.

4. The method for identifying parameters of a permanent magnet synchronous motor according to claim 3, characterized in that: Generate the initial population based on the Logistic map and set the current number of iterations t to 0; The initial position of each individual in the red-billed blue magpie population is calculated using the following formula: x t+1 =rx t (1-x t ) X i =lb j +x t+1 ·(ub j -lb j ),i=1,2,…,N, Where: x t is the chaotic vector of the tth iteration, x t+1 is the chaotic vector of the t+1th iteration, r is the mapping parameter; X i is the position of the i-th red-billed blue magpie in the search space, N is the number of populations, ub j and lb j They are the upper and lower limits of the search space, that is, the value range of the parameters to be identified.

5. The method for identifying parameters of a permanent magnet synchronous motor according to claim 4, characterized in that: The initial position of the red-billed blue magpie is obtained, and the estimated values ​​of the parameters to be identified are corrected accordingly. The motor parameter identification is carried out using the improved red-billed blue magpie optimization algorithm, which mainly includes three stages: finding food, attacking prey and preserving food. The food search phase: The mathematical model of food search is divided into two cases: small groups searching for food and large groups searching for food. The following is the corresponding mathematical formula: Where t represents the current iteration number, X i (t+1) represents the position of the i-th new search agent, p represents the number of hunts in a small group of red-billed blue magpies, q represents the number of hunts in a large group, X m represents the randomly selected mth individual, X rs (t) represents the randomly selected search agent in the current iteration, and Rand represents the variable used to generate random numbers. At this stage, the selection numbers p and q were improved. They were originally generated by uniform random distribution. This method may limit the range of variation of the selection number in each iteration and cannot adapt well to the scenario of motor parameter identification. A method of randomly generating the selection numbers p and q using normal distribution was introduced, which is suitable for application scenarios such as motor parameter identification that require precise control of the selection number. The randomness and distribution of the selection number can promote the exploration of diversity of individuals in the solution space, help to find better parameter identification solutions, and improve the convergence speed and stability of the algorithm. Attacking prey phase: In the attack phase, a tournament selection mechanism is introduced. The algorithm selects the best individual from the entire population through the tournament selection mechanism, and then decides whether to adopt a small group attack strategy or a group attack strategy to update the position of each individual based on the set probability Epsilon. Specifically, each individual first participates in a tournament, where individuals are randomly selected and compared based on fitness, and the winner will be used for subsequent updates. This process not only increases the diversity of the population, but also helps the algorithm to jump out of the local optimum and improve the global search capability. The formula for attacking prey can be summarized as: Among them: Epsilon is the probability of determining the attack strategy, X i is the current individual position, X food is the prey position, CF is the adjustment factor, where X p,mean and X q,mean is the average position recalculated based on the winner's position, X p,mean is the average value of p individual positions randomly selected from a normal distribution in a small population, p is a random number between 2 and 5, X q,mean is the average value of q individual positions randomly selected from a normal distribution in a large population, q is a random number between 10 and N, and rand is a random vector. Food preservation phase: In the Red-billed Blue Magpie Optimization (RBMO) algorithm, the formula of the food preservation phase is used to simulate the behavior of the Red-billed Blue Magpie storing food for future use. The purpose of this phase is to retain the information of the solution to help individuals find the global optimal solution during the optimization process. The following is the mathematical model of the food preservation phase: Where: X i (t+1) represents the position of individual i at time t+1; fitness old is the fitness value of the individual before updating its position; new is the fitness value of the individual after updating the position; X food (t) is the location of the food at time t, which is the optimal solution currently found.

6. The method for identifying parameters of a permanent magnet synchronous motor according to claim 5, characterized in that: The error between the actual output value and the theoretical output value is corrected according to the fitness function. When the maximum number of iterations is reached, the iteration is stopped and the optimal identification parameters are output.

Images