Surface-mounted permanent magnet synchronous motor multi-parameter identification method based on parameter self-adjusting PID search algorithm

Abstract

The invention discloses a surface-mounted permanent magnet synchronous motor multi-parameter identification method based on a parameter self-adjusting PID search algorithm, and the method comprises the steps: comparing an actual model of a permanent magnet synchronous motor with a parameter identification model, and constructing a fitness function to measure the difference between the actual model and the parameter identification model; a fitness value corresponding to a to-be-identified parameter is calculated by using a parameter self-adjustment-based PID search algorithm fitness function so as to perform iterative optimization, chaotic mapping is introduced in population initialization update, population diversity is increased, convergence speed is accelerated, and the robustness of the algorithm is improved according to changes of population search behaviors. A self-tuning PID parameter based on an exponential function is provided, rapid response to error changes is improved, and rapid convergence of searching early population is promoted. The PID search algorithm based on parameter self-adjustment can quickly and accurately identify the stator resistance, inductance and flux linkage parameters of the permanent magnet synchronous motor, and has obvious advantages in convergence rate and calculation amount.

Description

Technical Field

[0001] The present invention relates to the field of permanent magnet synchronous motors, and particularly relates to a multi-parameter identification method for surface-mounted permanent magnet synchronous motors based on a PID search algorithm with parameter self-regulation. Background Art

[0002] Due to its advantages such as high efficiency, rapid response, and compact structure, permanent magnet synchronous motors are widely used in industrial automation, electric vehicles, household appliances and other fields. However, the precise control of permanent magnet synchronous motors depends on the accurate identification of their parameters, such as stator resistance, stator inductance, and magnetic flux. In the actual working environment of permanent magnet synchronous motors, factors such as changes in motor temperature, load disturbances, and harmonic interference will cause changes in the motor parameters of permanent magnet synchronous motors, which will in turn lead to a decrease in the working efficiency of permanent magnet synchronous motors. Therefore, in the process of permanent magnet synchronous motor control, it is crucial to accurately identify the motor parameters.

[0003] Existing parameter identification methods are mainly divided into traditional methods (such as recursive least squares method, extended Kalman filter) and methods based on artificial intelligence optimization (such as particle swarm optimization, grey wolf algorithm). Traditional methods have problems such as sensitivity to noise, complex calculations, and parameter coupling; although existing intelligent algorithms have improved the convergence speed, they are prone to falling into local optima and do not fully consider the influence of inverter nonlinear characteristics on the identification results.

[0004] The PID search algorithm is a new intelligent algorithm proposed by Gao Yuansheng in 2023. This algorithm solves multi-dimensional problems by simulating the control of incremental PID. Due to its simple structure and high convergence accuracy, it can be used for parameter identification of permanent magnet synchronous motors. However, the PID search algorithm has the same problems as other meta-heuristic intelligent algorithms, that is, the convergence speed is slow and the local optimization ability is limited, which affects the accuracy of motor parameter identification and is prone to falling into local optima. Summary of the Invention

[0005] Aiming at the above-mentioned technical deficiencies, the purpose of the present invention is to provide a multi-parameter identification method for surface-mounted permanent magnet synchronous motors based on a PID search algorithm with parameter self-regulation.

[0006] To solve the above technical problems, the present invention adopts the following technical solutions:

[0007] The present invention provides a multi-parameter identification method for surface-mounted permanent magnet synchronous motors based on a PID search algorithm with parameter self-regulation, including the following steps:

[0008] (1) Construct a parameter identification model of the permanent magnet synchronous motor, including constructing the permanent magnet synchronous motor in the direct-axis component current i d = 0 and the direct-axis component current i dIntegrate the discrete voltage equations under two control strategies of ≠0 to obtain a fourth-order full-rank parameter identification model;

[0009] (2) Based on the difference between the identification values of the synchronous motor parameters under the current parameters to be identified and the parameter identification model described in step (1), construct a fitness function:

[0010]

[0011] Where, and are the parameters to be estimated of the stator resistance, the d-axis inductance, the q-axis inductance, and the permanent magnet flux linkage respectively; w1, w2, w3, w4 are weight coefficients, and Denote the actual value of the d-axis k-th voltage under the i d =0 control strategy as The identification value obtained according to the parameter identification model is The actual value of the q-axis k-th voltage is The identification value obtained according to the parameter identification model is Denote the actual value of the d-axis k-th voltage under the i d ≠0 control strategy as The identification value obtained according to the parameter identification model is The actual value of the q-axis k-th voltage is The identification value obtained according to the parameter identification model is

[0012] (3) Use a chaotic map to initialize the position of each individual in the population of the PID search algorithm in the solution space of the parameters to be identified; in each iteration, use the position of each individual as the parameter to be identified and calculate the fitness value of the individual according to the fitness function in step (2), and use the individual with the best fitness value in the current iteration as the given value to guide the position update of other individuals towards the best individual; the position update strategy for the individuals in the current iteration includes: dynamically adjusting the PID search algorithm coefficients to balance the search ability of the algorithm in real time according to the number of current iterations, and using the position of the individual with the best fitness in the current iteration as the global optimal reference to guide other individuals to approach it;

[0013] (4) Based on the PID search algorithm in step (3), iteratively update the position of the population of the PID search algorithm until the iteration termination condition is reached, output the position of the best individual as the identification result of the parameter to be identified, and construct an expression of the PID search algorithm based on parameter self-regulation. The PID adjustment output variable expression is:

[0014] Δu(t) = K p ·r2·Δe(t) + K i ·r3·e(t) + Kd ·r4·[e(t) - 2Δe(t - 1) + Δe(t - 2)]

[0015] where r2, r3, and r4 are random vectors with values ranging from 0 to 1; K p , K i , K d are the adjustment coefficients of the proportional term, integral term, and derivative term respectively; t represents the number of iterations; e(t) represents the population deviation at the t-th iteration, Δe(t) represents the change in population deviation between the t-th iteration and the (t - 1)-th iteration, Δe(t - 1) represents the population deviation generated at the previous iteration when the current iteration number is t, and Δe(t - 2) represents the population deviation generated at the two previous iterations when the current iteration number is t.

[0016] Among them, according to K p mainly affects the PSA search step size, enhances the global search ability in the early stage, and avoids falling into local optima; K i affects the cumulative correction of the search direction, enables the algorithm to converge more precisely in the later stage, and improves the local search accuracy; K d adjusts the search trend, avoids violent fluctuations, and improves the convergence speed. The self-adjusting parameter expression is set as:

[0017]

[0018] where e is a natural constant; ε is the adjustment rate coefficient and ε = 0.1, K p,initial , K i,initial , K d，initial represent the initial values of the adjustment coefficients of the proportional term, integral term, and derivative term respectively, and are set as 1, 1, 3;

[0019] Since the optimal individual value changes with each iteration, when the number of iterations is t, the population deviation expressions corresponding to the current and previous iterations are:

[0020]

[0021] where x best (t) and x best (t - 1) represent the best populations corresponding to the minimum fitness values in the t-th and (t - 1)-th iterations respectively, x(t - 1) represents the population in the (t - 1)-th iteration process, and e(t - 1) represents the population deviation generated at the (t - 1)-th iteration.

[0022] Preferably, in step (3), a chaotic mapping is used to initialize the position of each individual in the PID search algorithm population, including: the Logistic chaotic mapping expression y i = λ1y (i-1)(1 - y (i-1) ), when \(i = 1\), the initial value \(y_0\) is a random number between \([0, 1]\), and \(\lambda_1\in(0, 4]\); the initial population expression of the PID search algorithm is \(x\) ij \(=(u\) j \(-l\) j )r_1 + l\) j , (\(i = 1, 2, \cdots, n\); \(j = 1, 2, \cdots, h\)), \(r_1\) is a random number between \(0\) and \(1\); replacing the random number \(r_1\) in the initial population expression of the PID search algorithm with the mathematical expression of the Logistic mapping, the individual position representation of the initialized population introduced with the chaotic mapping is obtained as: \(x\) ij \(=(u\) j \(-l\) j )y\) i +l\) j ; in the formula, \(x\) ij represents the \(j\)-th dimension of the \(i\)-th individual; \(u\) j and \(l\) j are the upper and lower bounds of the \(j\)-th parameter respectively.

[0023] Preferably, the parameter identification model of the fourth-order full-rank permanent magnet synchronous motor constructed in step (1) is:

[0024]

[0025] where: \(i\) q0 (k)\) is the \(k\)-th current on the \(q\)-axis under the \(i\) d \( = 0\) control strategy, \(i\) q1 (k)\) is the \(k\)-th current on the \(q\)-axis under the \(i\) d \(\neq 0\) control strategy, \(i\) d1 (k)\) is the \(k\)-th current on the \(d\)-axis under the \(i\) d \(\neq 0\) control strategy, \(\omega\) e0 (k)\) is the \(k\)-th rotor angular velocity under the \(i\) d \( = 0\) control strategy, \(\omega\) e1 (k)\) is the \(k\)-th rotor angular velocity under the \(i\) d \(\neq 0\) control strategy.

[0026] Preferably, in step (3), the individual with the optimal fitness value in the current iteration is used as the given value to guide the position update of other individuals towards the optimal individual, specifically including:

[0027] Starting from the second iteration, when the fitness value of the individual with the optimal fitness value in the current iteration is greater than the fitness value of the optimal solution determined in the previous time, the optimal solution of the previous time is retained; otherwise, the individual with the optimal fitness value in the current iteration batch is used as the updated optimal solution; among them, the smaller the fitness value, the better the fitness value.

[0028] Preferably, in step (3), the position of the individual with the optimal fitness in the current iteration is used as the global optimal reference to guide other individuals to approach it. Specifically, it includes:

[0029] The position of the individual in the improved initial population is expressed as: x ij =(u j -l j )y i +l j ; i = 1, 2,..., n; j = 1, 2,..., h.

[0030] Among them, x ij represents the j-th dimension of the i-th individual; u j and l j are the upper and lower bounds of the j-th parameter respectively; n represents the population size; h is the dimension of the solution, that is, the number of estimated parameters, and the value is 4.

[0031] Preferably, in step (4), the population update formula is x(t + 1) = x(t) + ηΔu(t) + (1 - η)o(t); where η represents the adjustment factor, expressed as η = r6cos(t / T), where r6 is a random vector between 0 and 1; T represents the maximum number of iterations; o(t) is the zero-output adjustment factor, defined as: o(t) = (cos(1 - t / T) + λr5·L)·e(t), r5 is a random vector between 0 and 1, where λ is the adjustment coefficient, calculated by λ = [ln(T - t + 2) / ln(T)] 2 obtained; L is the random search path, and its expression is:

[0032]

[0033] In the formula, L is the L′evy flight path, and u and v satisfy the normal distribution,

[0034] σ u , σ v The expression of is:

[0035] σ v = 1

[0036] Among them, Г is a gamma function; β1 is the adjustment factor, set to 1.5.

[0037] The beneficial effects of the present invention are as follows:

[0038] (1) The algorithm of the present invention introduces a chaotic mapping in the population initialization update, increasing the population diversity and accelerating the convergence speed.

[0039] (2) According to the changes in group search behavior, the present invention proposes a self-tuning PID parameter based on an exponential function, which improves the rapid response to error changes and promotes the rapid convergence of the population in the early stage of search.

[0040] (3) The PID search algorithm based on parameter self-regulation proposed by the present invention can quickly and accurately identify the stator resistance, inductance and flux linkage parameters of a permanent magnet synchronous motor, and has obvious advantages in terms of convergence speed and computational complexity.

[0041] (4) The present invention adds a zero-output adjustment factor to the algorithm, which can prevent the algorithm from falling into local optimum. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0043] Figure 1 It is the overall control system block diagram of the present invention;

[0044] Figure 2 It is the schematic diagram of the parameter identification method of a permanent magnet synchronous motor based on the PID search algorithm with parameter self-regulation of the present invention;

[0045] Figure 3 It is the flowchart of the PID search algorithm based on parameter self-regulation;

[0046] Figure 4 It is the comparison diagram of fixed and self-regulated PID parameters of the present invention;

[0047] Figure 5(a) is the fitness function curve of the simulation of the present invention;

[0048] Figure 5(b) is the stator resistance identification curve of the simulation of the present invention;

[0049] Figure 5(c) is the stator inductance identification curve of the simulation of the present invention;

[0050] Figure 5(d) is the stator flux linkage identification curve of the simulation of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] As Figures 1 to 5(d) shown, this embodiment provides a multi-parameter identification method for a surface-mounted permanent magnet synchronous motor based on a PID search algorithm with parameter self-regulation.

[0053] This embodiment discloses a multi-parameter identification method for a surface-mounted permanent magnet synchronous motor based on a PID search algorithm with parameter self-regulation. According to Figure 1 the shown overall control system block diagram, the overall control system block diagram of this embodiment does not consider non-linear compensation. This embodiment controls the permanent magnet synchronous motor based on a speed-current double closed-loop control system. The three-phase stator currents i a 、i b 、i c of the permanent magnet synchronous motor are measured by current sensors, and the d-axis and q-axis currents i d and i q in the rotating coordinate system are obtained through Clarke and Park transformations. The real-time speed v and the rotor electrical angle θ are obtained through an encoder, and then the actual value n of the rotational speed and the rotor angular velocity ω e can be obtained by conversion. This application adopts the vector control principle, inputs the difference between the given value n ref of the rotational speed and the actual value n of the rotational speed into a PI controller to obtain the given value i qref of the q-axis current, inputs the difference between the given value i qref of the q-axis current and the q-axis current i q through a PI controller to obtain the q-axis voltage u q ; inputs the difference between the given value i dref of the d-axis current and the d-axis current i d through a PI controller to obtain the d-axis voltage u d ; after performing an inverse Park transformation on the q-axis voltage u q and the d-axis voltage u d to obtain the stator voltages u α and u β in the two-phase stationary coordinate system, they are used as the inputs of voltage space vector modulation to generate the switching signals of the inverter, and finally drive the permanent magnet synchronous motor.

[0054] In the above control process, the q-axis current i q 、the d-axis current i d 、the q-axis voltage u q 、the d-axis voltage u d and the rotor angular velocity ωe As the input of the permanent magnet synchronous motor parameter identification method in this embodiment, the parameters to be identified can be iteratively solved based on the improved PID search algorithm. For the functional schematic diagram of the permanent magnet synchronous motor parameter identification method, refer to Figure 2 the schematic diagram; the flowchart of the PID search algorithm based on parameter self-regulation is as Figure 3 shown. The specific implementation is as follows:

[0055] Step 1: Construct a parameter identification model for the permanent magnet synchronous motor, where the parameter identification model reflects the relationship between multiple parameters to be identified and the synchronous motor parameters;

[0056] Perform a coordinate transformation on the motor equation of the permanent magnet synchronous motor to establish the equation of the permanent magnet synchronous motor in the dq coordinate system as:

[0057]

[0058] According to the vector control principle of the permanent magnet synchronous motor as Figure 1 shown, when the permanent magnet synchronous motor is in a stable operating state, the current differential components of the d-axis and q-axis are equal to 0. Then, by setting i dref = 0, the discrete voltage equation of the permanent magnet synchronous motor can be obtained under the control strategy of i d = 0 as:

[0059]

[0060] Among them, in order to facilitate the distinction of different control strategies, subscripts are added to each parameter for distinction; i q0 (k) is the kth current of the q-axis under the control strategy of i d = 0, ω e0 (k) is the kth rotor angular velocity under the control strategy of i d = 0; U d0 (k) is the actual value of the kth voltage of the d-axis under the control strategy of i d = 0, which is U d0 (k), and U q0 (k) is the actual value of the kth voltage of the q-axis under the control strategy of i d = 0.

[0061] The parameters to be identified in this embodiment include the stator resistance R s , the permanent magnet flux linkage ψ f , the d-axis inductance L d , and the q-axis inductance L q a total of four. The method of injecting a negative-sequence weak magnetic current into the d-axis is used to solve the under-rank problem, that is, by setting i dref to take a negative-sequence weak magnetic current, the discrete voltage equation of the permanent magnet synchronous motor can be obtained under the control strategy of i d ≠0 as:

[0062]

[0063] wherein, i q (k) is the k-th q-axis current under the i d ≠0 control strategy, and i d (k) is the k-th d-axis current under the i d ≠0 control strategy, and ω e (k) is the k-th rotor angular velocity under the i d ≠0 control strategy.

[0064] Integrate the discrete voltage equations of the permanent magnet synchronous motor under the two control strategies of i d =0 and i d ≠0 to obtain a fourth-order full-rank parameter identification model as follows:

[0065]

[0066] In the formula, and are parameters to be estimated.

[0067] Use the above fourth-order full-rank parameter identification model to obtain four parameters to be identified: stator resistance R s , permanent magnet flux linkage ψ f , d-axis inductance L d and q-axis inductance L q .

[0068] Step 2: Construct a fitness function according to the difference between the actual model of the synchronous motor parameters and the parameter identification model described in Step 1.

[0069] Transform the parameter identification problem of the permanent magnet synchronous motor into an optimization problem. By comparing the existing actual model of the synchronous motor and the established parameter identification model, construct a suitable fitness function to characterize the difference between the two. The smaller the fitness value, the smaller the difference between the two, that is, the closer the identified values of the synchronous motor parameters obtained by the parameter identification model are to the actual values, indicating that the parameters to be identified are more accurate. Use the PID search algorithm to perform iterative optimization based on this fitness function, so as to obtain the smallest and thus optimal parameters to be identified.

[0070] During the control operation of the permanent magnet synchronous motor, the collected synchronous motor voltage parameters include: i d The actual value of the k-th d-axis voltage under the =0 control strategy is written as u d0 (k), i d The actual value of the k-th q-axis voltage under the =0 control strategy is written as u q0 (k), i dThe actual value of the k-th voltage on the d-axis under the ≠0 control strategy is written as u d1 (k), i d The actual value of the k-th voltage on the q-axis under the ≠0 control strategy is written as u q1 (k).

[0071] Under the current stator resistance R s , permanent magnet flux linkage ψ f , d-axis inductance L d and q-axis inductance L q , combined with the input rotor angular velocity and dq-axis currents, the identified values of the synchronous motor voltage parameters are obtained according to the parameter identification model, including: i d The identified value of the k-th voltage on the d-axis under the i =0 control strategy is written as d The identified value of the k-th voltage on the q-axis under the i =0 control strategy is written as d The identified value of the k-th voltage on the d-axis under the ≠0 control strategy is written as i d The identified value of the k-th voltage on the q-axis under the ≠0 control strategy is written as

[0072] By comparing the actual model and the identified model of the synchronous motor, the following fitness function is constructed:

[0073]

[0074] Step 3, in the solution space of the parameters to be identified, use the chaotic mapping to initialize the positions of each individual in the population of the PID search algorithm; in each iteration, use the position of each individual as the parameter to be identified and calculate the fitness value of the individual according to the fitness function in step (2), and use the individual with the best fitness value in the current iteration as the given value to guide the other individuals to update their positions towards the best individual. Starting from the second iteration, when the fitness value of the individual with the best fitness value in the current iteration is greater than the fitness value of the best solution determined in the previous time, retain the best solution in the previous time; otherwise, use the individual with the best fitness value in the current iteration batch as the updated best solution; among them, the smaller the fitness value, the better the fitness value; the position update strategy for the individuals in the current iteration includes: dynamically adjusting the PID search algorithm coefficients to balance the search ability of the algorithm in real time according to the number of current iterations, and using the position of the individual with the best fitness in the current iteration as the global optimal reference to guide the other individuals to approach it.

[0075] The traditional PID search algorithm randomly initializes the initial positions of each individual in the population, resulting in the lack of diversity and uneven distribution in space of the PID search algorithm population during initialization, and thus leading to a slow initial convergence speed of the algorithm.

[0076] In this embodiment, the chaotic mapping is used to initialize the positions of individuals in the population of the PID search algorithm. The Logistic chaotic mapping is used to initialize the position of any individual in the population of the PID search algorithm. The Logistic chaotic mapping is also known as the unimodal mapping, which is simple to implement and generates a uniformly distributed sequence.

[0077] It includes: the Logistic chaotic mapping expression y i = λ1y (i-1) (1 - y (i-1) ), when i = 1, the initial value y0 is a random number between [0, 1], and λ1 ∈ (0, 4]; the initial population expression of the PID search algorithm is x ij = (u j - l j )r1 + l j , (i = 1, 2,..., n; j = 1, 2,..., h), r1 is a random number between 0 and 1; replacing the random number r1 in the initial population expression of the PID search algorithm with the mathematical expression of the Logistic mapping, the expression for representing the individual position of the initialized population by introducing the chaotic mapping is obtained as: x ij = (u j - l j )y i + l j ; in the formula, x ij represents the j-th dimension of the i-th individual; u j and l j are respectively the upper and lower bounds of the j-th parameter.

[0078] Step 4, based on the PID search algorithm in Step 3, iteratively update the positions of the population of the PID search algorithm until the iteration termination condition is reached, and output the position of the optimal individual as the identification result of the parameter to be identified, and construct the PID search algorithm expression based on parameter self-regulation. The PID regulation output variable expression is:

[0079] Δu(t) = K p ·r2·Δe(t) + K i ·r3·e(t) + K d ·r4·[e(t) - 2Δe(t - 1) + Δe(t - 2)]

[0080] In the formula, r2, r3, and r4 are random vectors with a value range of 0 to 1; K p , K i , K dThey are the adjustment coefficients of the proportional term, integral term, and differential term respectively; t represents the number of iterations; e(t) represents the population deviation at the t-th iteration, Δe(t) represents the change in the population deviation between the t-th iteration and the (t - 1)-th iteration, Δe(t - 1) represents the population deviation generated at the previous iteration when the current number of iterations is t, and Δe(t - 2) represents the population deviation generated at the two previous iterations when the current number of iterations is t.

[0081] Among them, according to K p mainly affects the PSA search step size, enhances the global search ability in the early stage, and avoids falling into local optima; K i affects the cumulative correction of the search direction, enables the algorithm to converge more precisely in the later stage, and improves the local search accuracy; K d adjusts the search trend, avoids violent fluctuations, and improves the convergence speed. The self-adjusting parameter expression is set as:

[0082]

[0083] In the formula, e is a natural constant; ε is the adjustment rate coefficient and ε = 0.1, K p,initial , K i,initial , K d，initial represent the initial values of the adjustment coefficients of the proportional term, integral term, and differential term respectively, which are set to 1, 1, 3;

[0084] Since the optimal individual value changes with each iteration, when the number of iterations is t, the expressions for the population deviation generated in the current and previous iterations are:

[0085]

[0086] In the formula, x best (t) and x best (t - 1) represent the best populations corresponding to the minimum fitness values in the t-th and (t - 1)-th iterations respectively, x(t - 1) represents the population in the (t - 1)-th iteration process, and e(t - 1) represents the population deviation generated at the (t - 1)-th iteration.

[0087] The population update formula is x(t + 1) = x(t) + ηΔu(t) + (1 - η)o(t); where η represents the adjustment factor, which is expressed as η = r6cos(t / T), in the formula r6 is a random vector between 0 and 1; o(t) is the zero output adjustment factor, which is defined as: o(t) = (cos(1 - t / T) + λr5·L)·e(t), where λ is the adjustment coefficient, calculated by λ = [ln(T - t + 2) / ln(T)] 2 obtained, T represents the maximum number of iterations; r5 is a random vector between 0 and 1; L is the random search path, and its expression is:

[0088]

[0089] In the formula, L is the Lévy flight path, and u and v follow a normal distribution.

[0090] σ u , σ v The expression of... is:

[0091] s v = 1

[0092] where Γ is a gamma function; β1 is an adjustment factor set to 1.5.

[0093] To verify the effectiveness and feasibility of the parameter identification of the permanent magnet synchronous motor based on the parameter self - regulating PID search algorithm proposed in this paper, a surface - mounted permanent magnet synchronous motor identification system based on the parameter self - regulating PID search algorithm was built through the MATLAB software platform.

[0094] The parameters of the permanent magnet synchronous motor are as follows:

[0095] parameter numerical value parameter numerical value <![CDATA[Rated stator resistance R s > 0.22 Ω <![CDATA[Number of pole pairs P n > 5 <![CDATA[Rated d-axis inductance L d > 0.225 mH moment of inertia J <![CDATA[0.000028kg·m 2 > <![CDATA[Rated q-axis inductance L q > 0.225 mH <![CDATA[Rated torque T N > 0.64 Nm <![CDATA[Nominal permanent magnet flux linkage ψ f > 0.009824 Wb <![CDATA[Switching frequency f s > 10 kHz <![CDATA[Bus voltage u dc > 48V

[0096] In the algorithm, the sampling period is 0.1 ms, the number of data samples N is 500, the population size is set to 50, the maximum number of iterations is set to 200, R s , L d , L q and ψ f The identification parameter ranges of... are set to [0, 5], [0, 1], [0, 1], [0, 1].

[0097] The identification results of the algorithm are shown in Figures 5(a), 5(b), 5(c), and 5(d).

[0098] The fitness convergence iteration curve of the algorithm is shown in Figure 5(a). It can be seen that the convergence speed of the parameter self - regulating PID search algorithm is relatively fast and the accuracy is relatively high.

[0099] The identification result of the stator resistance is shown in Figure 5(b). It can be seen that the resistance identification value of the parameter self - regulating PID search algorithm is relatively close to the nominal value.

[0100] The inductance L of the surface - mounted permanent magnet synchronous motor d The identification result is shown in Figure 5(c). It can be seen that the parameter self - regulating PID search algorithm can converge quickly in the early stage of iteration and has a high identification accuracy.

[0101] The flux linkage identification results of the surface-mounted permanent magnet synchronous motor are shown in Fig. 5(d). It can be seen that the PID search algorithm based on parameter self-regulation can quickly identify the flux linkage with a small identification error.

[0102] In summary, the PID search algorithm based on parameter self-regulation has a fast convergence speed, high identification accuracy, and good stability and small fluctuations during the iteration process.

[0103] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. A multi-parameter identification method for surface-mounted permanent magnet synchronous motors based on a PID search algorithm with parameter self-regulation, characterized in that, Including the following steps: (1)Construct a parameter identification model for a permanent magnet synchronous motor, including constructing a discrete voltage equation under two control strategies of the permanent magnet synchronous motor, i.e., the direct-axis component current i d = 0 and the direct-axis component current i d ≠ 0, and integrating them to obtain a fourth-order full-rank parameter identification model; (2) Based on the difference between the identification values of the synchronous motor parameters of the parameter identification model described in step (1) under the current parameter to be identified, construct a fitness function: Among them, and are the parameters to be estimated for the stator resistance, the d-axis inductance, the q-axis inductance, and the permanent magnet flux linkage respectively; w1, w2, w3, w4 are the weight coefficients, and Denote the actual value of the k-th d-axis voltage under the i d =0 control strategy as The identification value obtained according to the parameter identification model is The actual value of the k-th q-axis voltage is The identification value obtained according to the parameter identification model is Denote the actual value of the k-th d-axis voltage under the i d ≠0 control strategy as The identification value obtained according to the parameter identification model is The actual value of the k-th q-axis voltage is The identification value obtained according to the parameter identification model is (3) Initialize the positions of each individual in the PID search algorithm population using a chaotic map within the solution space of the parameter to be identified; in each iteration, use the position of each individual as the parameter to be identified and calculate the fitness value of the individual according to the fitness function in step (2), and use the individual with the optimal fitness value in the current iteration as the given value to guide the position update of other individuals towards the optimal individual; the position update strategy for the individuals in the current iteration includes: dynamically adjusting the PID search algorithm coefficients to balance the search ability of the algorithm in real time according to the number of current iterations, and using the position of the individual with the optimal fitness value in the current iteration as the global optimal reference to guide other individuals to approach it; (4) Based on the PID search algorithm in step (3), iteratively update the positions of the PID search algorithm population until the iteration termination condition is reached, and output the position of the optimal individual as the identification result of the parameter to be identified, and construct an expression of the PID search algorithm based on parameter self-regulation. The PID regulation output variable expression is: Δu(t) = K p ·r2·Δe(t) + K i ·r3·e(t) + K d ·r4·[e(t) - 2Δe(t - 1) + Δe(t - 2)] where r2, r3, and r4 are random vectors with values ranging from 0 to 1; K p , K i , K d are the adjustment coefficients of the proportional term, integral term, and derivative term, respectively; t represents the number of iterations; e(t) represents the population deviation at the t-th iteration, Δe(t) represents the change in the population deviation between the t-th iteration and the (t - 1)-th iteration, Δe(t - 1) represents the population deviation generated at the previous iteration when the current iteration number is t, and Δe(t - 2) represents the population deviation generated at the two previous iterations when the current iteration number is t; Among them, according to K p mainly affects the PSA search step size, enhances the global search ability in the early stage, and avoids falling into local optima; K i affects the cumulative correction of the search direction, enables the algorithm to converge more precisely in the later stage, and improves the local search accuracy; K d adjusts the search trend, avoids violent fluctuations, improves the convergence speed, and sets the self-adjusting parameter expression as: where e is a natural constant; ε is the adjustment rate coefficient and ε = 0.1, K p,initial ,K i,initial ,K d，initial represent the initial values of the adjustment coefficients of the proportional term, integral term and derivative term respectively, which are set to 1, 1, 3 respectively; Since the optimal individual value changes in each iteration, when the number of iterations is t times, the population deviation expressions generated by the current and the previous iteration are: where x best (t) and x best (t - 1) represent the best populations corresponding to the minimum fitness values in the t-th and (t - 1)-th iterations respectively, x(t - 1) represents the population in the (t - 1)-th iteration process, and e(t - 1) represents the population deviation generated at the (t - 1)-th iteration.

2. The multi-parameter identification method of the surface-mounted permanent magnet synchronous motor based on the PID search algorithm with parameter self-regulation according to claim 1, characterized in that, In step (3), the chaotic mapping is used to initialize the positions of each individual in the population of the PID search algorithm, including: the Logistic chaotic mapping expression y i = λ1y (i-1) (1 - y (i-1) ). When i = 1, the initial value y0 is a random number between [0, 1], and λ1 ∈ (0, 4]; the initial population expression of the PID search algorithm is x ij = (u j - l j )r1 + l j , (i = 1, 2,..., n; j = 1, 2,..., h), r1 is a random number between 0 and 1; replace the random number r1 in the initial population expression of the PID search algorithm with the mathematical expression of the Logistic mapping, and the individual position representation of the initialized population introduced by the chaotic mapping is obtained as: x ij = (u j - l j )y i + l j ; in the formula, x ij represents the jth dimension of the ith individual; u j and l j are the upper and lower bounds of the jth parameter respectively.

3. The multi-parameter identification method of the surface-mounted permanent magnet synchronous motor based on the PID search algorithm with parameter self-regulation as claimed in claim 1, characterized in that, The parameter identification model of the fourth-order full-rank permanent magnet synchronous motor constructed in step (1) is: where: i q0 (k) is the k-th q-axis current under the i d =0 control strategy, and i q1 (k) is the k-th q-axis current under the i d ≠0 control strategy, and i d1 (k) is the k-th d-axis current under the i d ≠0 control strategy, ω e0 (k) is the k-th rotor angular velocity under the i d =0 control strategy, and ω e1 (k) is the k-th rotor angular velocity under the i d ≠0 control strategy.

4. The multi-parameter identification method of the surface-mounted permanent magnet synchronous motor based on the PID search algorithm with parameter self-regulation according to claim 1, characterized in that, In step (3), using the individual with the optimal fitness value in the current iteration as the given value to guide the position update of other individuals towards the optimal individual specifically includes: Starting from the second iteration, when the fitness value of the individual with the optimal fitness value in the current iteration is greater than the fitness value of the optimal solution determined in the previous time, retain the optimal solution in the previous time; otherwise, use the individual with the optimal fitness value in the current iteration batch as the updated optimal solution; among them, the smaller the fitness value, the better the fitness value represents.

5. The multi-parameter identification method of a surface-mounted permanent magnet synchronous motor based on a PID search algorithm with parameter self-regulation as claimed in claim 1, wherein, In step (3), using the position of the individual with the optimal fitness value in the current iteration as the global optimal reference to guide other individuals to approach it specifically includes: The individual positions of the initialized population are represented as: x ij =(u j -l j )y i +l j ; i = 1, 2, ..., n; j = 1, 2, ..., h; where x ij represents the j-th dimension of the i-th individual; u j and l j are the upper and lower bounds of the j-th parameter respectively; n represents the population size; h is the dimension of the solution, i.e., the number of estimated parameters, taking the value of 4.

6. The multi-parameter identification method of the surface-mounted permanent magnet synchronous motor based on the PID search algorithm with parameter self-regulation as claimed in claim 1, wherein, In step (4), the population update formula is x(t + 1) = x(t) + ηΔu(t) + (1 - η)o(t); where η represents the adjustment factor, expressed as η = r6cos(t / T), where r6 is a random vector between 0 and 1; T represents the maximum number of iterations; o(t) is the zero output adjustment factor, defined as: o(t) = (cos(1 - t / T) + λr5·L)·e(t), r5 is a random vector between 0 and 1, where λ is the adjustment coefficient, calculated by λ = [ln(T - t + 2) / ln(T)] 2 ; L is the random search path, and its expression is: where L is the Levy flight path, and u and v follow a normal distribution. σ u , σ v The expression for Among them, Γ is a gamma function, and β1 is an adjustment factor, which is set to 1.5.

Images