Permanent magnet synchronous motor control method based on adaptive flux weakening control

Abstract

The invention relates to a permanent magnet synchronous motor control method based on self-adaptive flux weakening control. Compared with the prior art, the defect that a traditional control system is difficult to meet nonlinear transformation of gains of high-speed flux weakening and low-speed flux weakening voltage feedback loops at the same time is overcome. The method comprises the following steps: designing a self-adaptive controller; constructing a rotating speed regulation factor; constructing a current angle regulation factor; and optimizing fuzzy control by adopting a particle swarm algorithm. According to the method, three non-independent influence factors are converted into two independent influence factors, namely the rotating speed and the current angle, so that the influence of the rotating speed change and the current change on the voltage feedback gain is eliminated, the linearization of the voltage feedback gain is realized, and the nonlinear problem of a voltage control loop in the voltage feedback field weakening of the permanent magnet synchronous motor is solved.

Description

Technical Field

[0001] The invention relates to the technical field of permanent magnet synchronous motors, and in particular to a permanent magnet synchronous motor control method based on adaptive magnetic field weakening control. Background Art

[0002] The DC motors, induction motors, permanent magnet synchronous motors, switched reluctance motors, etc. that have been used or have application prospects in electric vehicles include. Permanent magnet synchronous motors (PMSM) are widely used in electric vehicles, aerospace, industrial control and other industries due to their high power density, high efficiency and good control characteristics.

[0003] Due to the large speed regulation range required for automotive motors, the high-speed performance of the motor needs to be fully guaranteed. The quality of the motor's high-speed performance is determined by the motor's own structure and electromagnetic parameters on the one hand, and is also affected by the control algorithm on the other. As the speed increases, the back electromotive force of the permanent magnet synchronous motor also continues to increase, and the speed range is limited by the DC side voltage. In order to further expand the speed range of PMSM, a weak magnetic control strategy needs to be adopted. However, due to the nonlinear characteristics of the motor parameters and the limitations of voltage and current, it is difficult to significantly improve the weak magnetic performance of PMSM.

[0004] The mainstream flux weakening control strategy can be divided into feedforward control, feedback control and hybrid control according to the control type. Feedforward flux weakening control is based on an accurate mathematical model and calculates the optimal reference current for the required torque and flux at each speed based on the motor parameters and the actual bus voltage. Due to the large amount of calculation, this method usually performs some simplifications from nonlinear to linear. It is very sensitive to motor parameters and operating conditions and it is difficult to guarantee the final performance. The hybrid control method uses both the feedforward method to generate a reference signal and the feedback closed loop to compensate for the voltage distortion caused by changes in motor parameters and the inverter. Feedback flux weakening directly controls the voltage output by the inverter through appropriate closed-loop regulation. This method has the advantages of good robustness and strong parameter adaptability and is widely used. Summary of the invention

[0005] The purpose of the present invention is to solve the defect that the traditional control system is difficult to simultaneously meet the nonlinear transformation of the gain of the high-speed weakening magnetic field and the low-speed weakening magnetic field voltage feedback loop, and to provide a permanent magnet synchronous motor control method based on adaptive weakening magnetic field control to solve the above problem.

[0006] In order to achieve the above object, the technical solution of the present invention is as follows:

[0007] A permanent magnet synchronous motor control method based on adaptive magnetic field weakening control comprises the following steps:

[0008] 11) Design of adaptive controller: Design adaptive coefficients to transform three non-independent influencing factors into two independent influencing factors: speed and current angle;

[0009] 12) Constructing the speed adjustment factor: Using an inverse function to construct the speed adjustment factor to eliminate the influence of speed change on the voltage feedback gain;

[0010] 13) Construct current angle adjustment factor: Design a fuzzy controller and use fuzzy control to construct the current control angle adjustment factor to eliminate the influence of current change on voltage feedback gain;

[0011] 14) Use particle swarm algorithm to optimize fuzzy control: Offline optimization of the output proportional factor and fuzzy rule weight of the fuzzy controller is performed to eliminate the randomness of the parameters of the fuzzy controller during design;

[0012] The optimized fuzzy controller parameters are substituted into the actual fuzzy controller to realize the linearization of voltage feedback gain; when the motor enters the low-speed weak magnetic field area, the adaptive controller adjusts the voltage feedback loop gain, so that the sensitivity of current angle regulation increases and the voltage regulation speed is accelerated; under medium and high speed weak magnetic field conditions, the adaptive controller adaptively adjusts the integral coefficient to ensure the stability of current even under small voltage margin.

[0013] The design of the adaptive controller includes the following steps:

[0014] 21) For each steady-state operating point, the gain of the voltage feedback loop is set as follows:

[0015]

[0016] Among them, ω e is the motor electrical angular velocity, I s is the motor stator current amplitude, β is the motor current control angle, U s is the motor stator voltage amplitude;

[0017] Since at a certain steady-state operating point, the voltage limit circle is determined as the speed is determined, when the current is determined, there must be a unique current control angle corresponding to it, so ω in the above formula is e ,I s and β are not independent in the weak magnetic region. If ω e With β and β as the two independent variables in the above equation, the gain of the voltage feedback loop is obviously nonlinear for these two independent variables;

[0018] 22) In order to obtain the control bandwidth and stability margin at each static operating point, the adaptive coefficient K is proposed adapt :

[0019] Based on the relatively slow change in speed and the rapid change in current, K adapt Decoupled into speed adjustment factor K ω and current control angle adjustment factor K β , the gain of the voltage feedback loop is kept almost constant by these two adjustment factors:

[0020] K adapt =K ω ·K β ,

[0021]

[0022] Among them, K adapt is the adaptive coefficient, K ω is the speed adjustment factor, K β is the current control angle adjustment factor, C is a constant;

[0023] 23) Design of adaptive controller: Adaptive coefficient is decoupled into speed adjustment factor K ω and current control angle adjustment factor K β , the current regulator and speed regulator output K respectively β and K ω , adjust the gain of the voltage feedback loop, and the output current control angle passes through the anti-saturation limiter to obtain the current control angle required for weak magnetic field.

[0024] The construction of the rotation speed adjustment factor comprises the following steps:

[0025] 31) Based on the fact that the maximum speed of the built-in permanent magnet synchronous motor in actual application can reach 3 to 4 times the base speed, assuming that the maximum speed is 3 times the base speed, the gain of the voltage control loop at the maximum speed is 9 times that at the base speed; in order to amplify the gain at low speed, the speed adjustment factor K ω The design is as follows:

[0026]

[0027] Where n and ω are the speed and angular velocity of the motor, n b and ω b It represents the rotation speed and angular velocity of the motor at the base speed, and k represents the ratio of the maximum speed to the base speed;

[0028] 32) When other conditions remain unchanged, the speed adjustment factor K ω To make the gain of the voltage control loop constant, when the motor operates below the base speed, that is, below the maximum speed at maximum torque output, the reduction in the denominator leads to K ω Increase; when the motor speed is lower than the basic speed, the limiter works so that the entire weak magnetic control loop has no effect on the motor and does not affect the stability of the control loop.

[0029] The construction of the current angle adjustment factor comprises the following steps:

[0030] 41) Fuzzy control is used to control the current angle adjustment factor. The input is the current control angle and its rate of change, and the output is the adjustment factor K. fuzzy , which describes the effect of the current control angle change on the gain of the weak magnetic feedback loop. At the same time, the current control angle adjustment factor K β satisfy:

[0031]

[0032] 42) Design of fuzzy controller:

[0033] The input of the fuzzy controller is the current control angle and its rate of change. The input is mapped from the physical domain to the fuzzy domain through the quantization factor to obtain a clear value. The clear value of the input is converted to the corresponding fuzzy subset and membership function through fuzzification. According to the fuzzy rules, the fuzzification is resolved to obtain the exact value, and then the output variable is mapped from the fuzzy domain to the physical domain by the proportional factor, and the output is the adjustment coefficient.

[0034] 43) Set the input and output of the fuzzy controller:

[0035] The input of the fuzzy controller is a two-dimensional vector consisting of β and Δβ, and the output is the adjustment factor K fuzzy , Δβ is the derivative of β, β and Δβ are respectively determined by the quantization factor K i1 and K i2 Mapped to [1,2] and [-1,1], K fuzzy Finally, it is mapped to [1,10] by the scale factor;

[0036] 44) Fuzzy processing of fuzzy controller:

[0037] β and K fuzzy It is described by seven natural language variables, as follows: negative large NB, negative medium NM, negative small NS, zero Z, positive small PS, positive medium PM, positive large PB;

[0038] Δβ is described by three linguistic variables: negative N, zero Z, positive P;

[0039] 45) Fuzzy controller defuzzification: The defuzzification process uses the centroid method to find the exact value of the output variable.

[0040]

[0041] In the formula, x i is the value of the horizontal axis, μ N (x i ) represents the membership function at x iThe value at

[0042] 46) Finally, adjust the coefficient K fuzzy Substitute into 41) to calculate the current control angle adjustment factor K β .

[0043] The optimization of the fuzzy controller by using the particle swarm algorithm comprises the following steps:

[0044] 51) The particle swarm algorithm is used to optimize the fuzzy controller. In order to make the designed fuzzy controller have higher control accuracy, the proportional coefficient of fuzzy control and the weight of fuzzy rules are used as the input of the particle swarm algorithm; the initial particle position vector is defined as

[0045]

[0046] Where p1 i (1×1) is a one-dimensional population, K o Represents the proportional factor of the fuzzy controller; p2 i (1×21) is a 21-dimensional population, w1,w2,…,w 21 They represent the weights of fuzzy rules respectively; i represents the member number in the population, and N represents the population size, that is, the number of particles in the population;

[0047] 52) The particle swarm algorithm selects the ITAE error of the speed as the objective function of the particle swarm. The effect of the control algorithm is measured by speed and torque, so the fitness function J ITAE Defined as:

[0048] J ITAE =∫t·|e|·dt,

[0049] In the formula, e is the speed error or torque error of the motor, and t is the action time;

[0050] 53) The particle swarm algorithm generates the initial particle position of the fuzzy controller.

[0051] All particles in the particle swarm adjust their speed and position for a finite number of iterations according to their individual extreme values ​​and the current global optimal solution. The specific iteration function is:

[0052]

[0053] In the formula, represents the d-dimensional velocity component of particle i in the k-th iteration, represents the d-dimensional position component of particle i in the k-th iteration, c1 and c2 are learning factors, which adjust the degree to which the particle draws on individual experience and group experience respectively; r1 and r2 are two random numbers that increase the randomness of the search; ω represents the weight factor for the particle to inherit its previous velocity; and are the individual optimum and global optimum obtained at the kth iteration, respectively. The individual optimum is a d-dimensional vector, and the global optimum is a value. represents the d-dimensional velocity component of particle i in the k+1th iteration, The optimal individual obtained in the k+1th iteration;

[0054] 54) In the iterative process, the increase of the inertia weight coefficient ω improves the global search ability of the particle swarm algorithm, and the decrease of the inertia weight coefficient enhances the algorithm's search accuracy for local extreme values. Based on this, the weight coefficient ω(k) is designed as:

[0055]

[0056] Among them, ω start is the initial inertia weight, ω end is the inertia weight when iterating to the maximum evolutionary generation, k is the number of iterations, T max is the maximum number of iterations;

[0057] 55) According to the iterative method of particle swarm algorithm, particles will move around so that the speed ITAE error gradually decreases. After multiple iterations, the particle swarm will tend to the global optimal solution as a whole;

[0058] 56) The optimized fuzzy controller parameters are obtained by particle swarm optimization algorithm, and the obtained parameters are substituted into the actual fuzzy controller to realize voltage feedback gain linearization;

[0059] When the motor enters the low-speed weak-magnetic field area, the adaptive controller adjusts the voltage feedback loop gain, which increases the sensitivity of the current angle regulation and speeds up the voltage regulation. Under medium- and high-speed weak-magnetic field conditions, the adaptive controller can adaptively adjust the integral coefficient to ensure current stability even under small voltage margins.

[0060] Beneficial Effects

[0061] Compared with the prior art, the present invention discloses a permanent magnet synchronous motor control method based on adaptive magnetic weakening control, which converts three non-independent influencing factors into two independent influencing factors, namely, rotational speed and current angle, so as to eliminate the influence of rotational speed change and current change on voltage feedback gain, realize voltage feedback gain linearization, and solve the nonlinear problem of voltage control loop in voltage feedback magnetic weakening of permanent magnet synchronous motor.

[0062] The present invention proposes an adaptive control strategy, which adopts an inverse function to construct a speed adjustment factor, adopts fuzzy control to construct a current control angle adjustment factor, realizes voltage feedback gain linearization, and adopts a particle swarm algorithm to optimize the proportional factor and fuzzy rule weight of the designed fuzzy controller, thereby enhancing the effect of the fuzzy controller and making the motor have better dynamic response characteristics in the low-speed weak magnetic field area; and can reduce the jitter of voltage and current in the high-speed weak magnetic field area. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 is a method sequence diagram of the present invention;

[0064] Figure 2 The structure diagram of the adaptive magnetic field weakening controller designed for the present invention;

[0065] Figure 3 It is the static gain diagram of the voltage control loop at the rated current amplitude;

[0066] Figure 4 Design the fuzzy controller structure diagram for the present invention;

[0067] Figure 5 The fuzzy rules of the current angle, the fuzzy rules of the current angle derivative, and the fuzzy rules of the adjustment coefficient of the fuzzy controller designed by the present invention are respectively displayed;

[0068] Figure 6 The fuzzy surface diagram of the fuzzy controller designed by the present invention;

[0069] Figure 7 The structure diagram of the particle swarm optimization fuzzy controller designed for the present invention;

[0070] Figure 8 This is a particle swarm optimization result diagram of the particle swarm algorithm of the present invention. DETAILED DESCRIPTION

[0071] In order to have a further understanding and recognition of the structural features and the effects achieved by the present invention, a preferred embodiment and accompanying drawings are used for detailed description as follows:

[0072] like Figure 1 As shown, a permanent magnet synchronous motor control method based on adaptive magnetic field weakening control comprises the following steps:

[0073] The first step is the design of the adaptive controller: design the adaptive coefficient to convert the three non-independent influencing factors into two independent influencing factors of speed and current angle, so as to linearize the voltage feedback gain.

[0074] The goal of magnetic field weakening control is to make the stator voltage amplitude less than the voltage limit by rationally planning the current trajectory. In order to study the specific performance of the magnetic field weakening link, the static gain of the transfer function of the stator voltage amplitude changing with the current space vector angle (also called the current control angle) is calculated under steady-state conditions at different speeds. Since the mechanical time constant of the motor is much larger than the electrical time constant, the speed change relative to the electrical signal in a short period of time can be ignored. The motor speed loop PI outputs the current vector magnitude, so the current vector magnitude can also be regarded as a constant relative to the voltage. The rate of change of inductance is relatively slow and can also be regarded as a constant. When the motor works in a magnetic field weakening condition, the current operating point is located on the voltage limit circle.

[0075] Using small signal analysis in the vicinity of a certain steady-state operating point, a small signal increment is given to the current control angle Then the electron voltage vector amplitude can be expressed as a steady-state value plus a small signal:

[0076]

[0077] Where U s is the electron voltage vector magnitude, Represent the steady-state voltage value and small signal increment of a certain working point, U lim is the voltage limit;

[0078] Linearizing the small signal term of the voltage amplitude yields:

[0079]

[0080] Expand the derivative term at the motor operating point to obtain:

[0081]

[0082] Since the stator voltage and current equation of PMSM in the dq coordinate system is as follows:

[0083]

[0084] In the formula, R s is the motor phase resistance, L d and L q are the d-axis and q-axis inductances of the permanent magnet synchronous motor, ψ f is the permanent magnet flux, ω e is the motor electrical angular velocity, i s is the stator current amplitude, and β is the current control angle.

[0085] We can get:

[0086]

[0087] That is, the steady-state gain of the current control angle to the voltage vector when the speed is steady. The resistance term can be ignored in the high-speed area, that is:

[0088]

[0089] Therefore, for each steady-state operating point, the gain of the voltage feedback loop can be expressed as:

[0090]

[0091] At a certain steady-state operating point, the voltage limit circle is determined as the speed is determined. When the current is determined, there must be a unique current control angle corresponding to it. In other words, ω in the above formula e , I s and β are not independent in the weak magnetic region. e With β and β as the two independent variables in the above equation, the gain of the voltage control loop is obviously nonlinear with respect to these two independent variables.

[0092] In order to obtain a more consistent control bandwidth and stability margin at each static operating point, an adaptive coefficient K is proposed adapt In fact, the change of speed is relatively slow, while the change of current is fast. For this reason, K adapt Decoupled into speed adjustment factor K ω and current control angle adjustment factor K β The gain of the voltage control loop can be kept almost constant by these two adjustment factors:

[0093] K adapt =K ω ·K β

[0094]

[0095] The adaptive magnetic field weakening controller designed by the present invention is as follows: Figure 2 , the adaptive coefficient is decoupled into the speed adjustment factor K ω and current control angle adjustment factor K β , the current regulator and speed regulator output K respectively β and K ω , adjust the gain of the voltage feedback loop, and the output current control angle passes through the anti-saturation limiter to obtain the current control angle required for weak magnetic field.

[0096] The second step is to construct the speed adjustment factor: use the inverse function to construct the speed adjustment factor to eliminate the influence of speed change on the voltage feedback gain.

[0097] For built-in permanent magnet synchronous motors, the maximum speed can reach 3 to 4 times the base speed in practical applications. Assuming that the maximum speed is 3 times the base speed, the gain of the voltage control loop at the maximum speed will be 9 times that at the base speed. In order to amplify the gain at low speed, the speed adjustment factor K ω The design is as follows:

[0098]

[0099] Where n b and ω b It represents the rotation speed and angular velocity of the motor at the base speed, and k represents the ratio of the maximum speed to the base speed.

[0100] When the motor operates below the base speed, that is, below the maximum speed at maximum torque output, due to the decrease in the denominator, K ω It will become very large, but when the motor speed is lower than the basic speed, the entire weak magnetic control loop will not have any effect on the motor due to the operation of the limiter, so it will not affect the stability of the control loop.

[0101] The third step is to construct the current angle adjustment factor: design a fuzzy controller and use fuzzy control to construct the current control angle adjustment factor to eliminate the influence of current change on the voltage feedback gain.

[0102] At a certain speed, I s and β are two related variables, so in order to s It is a function of β. It is necessary to solve a system of equations. The system of equations to be solved is a system of transcendental equations. The approximate curve is as follows: Figure 3 As shown, it is inappropriate to substitute the solution back into the gain equation for calculation. In addition, due to the sudden change of current in the transient process, the steady-state gain still cannot accurately adjust the gain of the voltage feedback loop in the transient process. Considering the above reasons, fuzzy control is used to solve these problems.

[0103] The fuzzy controller structure designed by the present invention is as follows Figure 4 As shown, the input is the current control angle and its rate of change, and the output is the adjustment coefficient K fuzzy , its design follows Figure 3 The gain shown varies with the current control angle adjustment factor K β Should meet:

[0104]

[0105] (1) Input and Output

[0106] Due to the variability of the external environment, it is generally hoped that the fuzzy domain remains unchanged to improve the adaptability of the fuzzy controller. In addition to transforming the domain and matching the front and back modules, the quantization factor and the proportional factor have a certain regulatory effect on the entire system, because its change is equivalent to scaling the measurement signal, which will directly affect the effect of the sampling signal on the system.

[0107] The input of the fuzzy controller is a two-dimensional vector consisting of β and Δβ (derivative of β). The output is the adjustment factor K fuzzy The physical range of β is [π / 2,π], and the physical range of Δβ is limited to [-5,5] by the limiter. β and Δβ are respectively determined by the quantization factor K i1 and K i2 Maps to [1,2] and [-1,1]. K fuzzy Finally it is mapped to [1,10] by the scaling factor.

[0108] (2) Fuzzification

[0109] In order to make the clear value participate in the fuzzy reasoning operation, the input clear value is converted to the corresponding fuzzy subset and membership function. fuzzy It is described by seven natural language variables: negative big (NB), negative middle (NM), negative small (NS), zero (Z), positive small (PS), positive middle (PM), positive big (PB). Δβ is described by three language variables: negative (N), zero (Z), positive (P).

[0110] As the current control angle increases, the gain of the voltage control loop shows stronger nonlinearity. A steep membership function will give the control a higher resolution, and the density of the membership function increases with the increase of the current control angle, and the control stability and accuracy of the key area will be improved. Specific membership function design: current angle fuzzy rule, current angle differential fuzzy rule, and adjustment coefficient fuzzy rule are as follows: Figure 5 .

[0111] (3) Fuzzy Inference Rules

[0112] The linguistic fuzzy rules are composed of "if...and..., then..." statements containing fuzzy relations, and each statement contains a control rule. For the convenience of observation, they can be converted into tabular fuzzy rules.

[0113] The fuzzy inference rules are shown in Table 1.

[0114] Table 1

[0115]

[0116] (4) Defuzzification

[0117] The defuzzification process uses the centroid method to find the exact value of the output variable.

[0118]

[0119] In the formula, x i is the value of the horizontal axis, μ N (x i ) represents the membership function at x i The final designed fuzzy surface is as follows Figure 6 .

[0120] Finally, the adjustment coefficient K fuzzy Substitute the above current control angle adjustment factor K β K is calculated in the formula β .

[0121] The fuzzy controller designed according to this rule can compensate for the gain change caused by the change of the current control angle, and because the current angle change rate is taken into account, a certain advance compensation can also be made for the dynamic response trend of the motor.

[0122] The fourth step is to use the particle swarm algorithm to optimize the fuzzy control: the output proportional factor and fuzzy rule weight of the fuzzy controller are optimized offline to eliminate the randomness of the parameters of the fuzzy controller during design.

[0123] Due to the influence of motor magnetic saturation, the gain of the voltage feedback loop is not always accurate under high current conditions. At the same time, the proportional factor and fuzzy rule of the fuzzy controller are not optimal. When the proportional factor is small, it takes a long time for the weak magnetic controller to reach a steady state; when the proportional factor is large, the control system is more sensitive to disturbances. Therefore, the present invention uses a particle swarm algorithm to optimize the fuzzy controller. The specific optimization process is as follows: Figure 7 shown.

[0124] The selection of optimization target and fitness function in particle swarm algorithm is the most critical step. In order to make the designed fuzzy controller have higher control accuracy, the proportional coefficient of fuzzy control and the weight of fuzzy rule are used as the input of particle swarm algorithm. The fuzzy controller designed above has two inputs and one output. The two inputs correspond to seven membership functions and three membership functions respectively, and a total of 21 fuzzy rules will be generated. Therefore, the initial particle position vector of the designed particle swarm is

[0125]

[0126] Where p1 i (1×1) is a one-dimensional population, K o Represents the proportional factor of the fuzzy controller; p2 i (1×21) is a 21-dimensional population, w1,w2,L,w 21They represent the weights of the fuzzy rules respectively; i represents the member number in the population; N represents the population size (the number of particles in the population), which is set to 20.

[0127] For motor weak magnetic field control, the main evaluation index can be the step response characteristics under weak magnetic field conditions. A good weak magnetic field control algorithm can make the motor still have a fast torque response capability under weak magnetic field conditions and quickly reach the target speed. The ITAE error of the speed is selected as the objective function of the particle swarm. The effect of the control algorithm can be measured by speed and torque, so the fitness function is defined as:

[0128] J ITAE =∫t·|e|·dt

[0129] In the formula, e is the speed error or torque error of the motor, and t is the action time;

[0130] The particle swarm algorithm generates the initial particle position of the fuzzy controller. Each particle has a random position in space when initialized, and all particles are relatively evenly distributed in space. Each particle searches for the optimal solution in space individually, records the optimal value it finds as the current individual extreme value, and compares the individual extreme value with the individual extreme values ​​of other particles in the entire particle swarm to find the optimal individual extreme value as the current global optimal solution of the entire particle swarm. All particles in the particle swarm adjust their speed and position according to the individual extreme value and the current global optimal solution. The specific iterative method is:

[0131]

[0132] In the formula, represents the d-dimensional velocity component of particle i in the k-th iteration; represents the d-dimensional position component of particle i in the kth iteration; c1 and c2 are learning factors, which adjust the degree to which the particle draws on individual experience and group experience respectively, and set c1=c2=1.49; r1 and r2 are two random numbers that increase the randomness of the search; ω represents the weight factor for the particle to inherit its previous speed; and are the individual optimum and global optimum obtained at the kth iteration, respectively. The individual optimum is a d-dimensional vector, and the global optimum is a value. represents the d-dimensional velocity component of particle i in the k+1th iteration, The optimal individual is obtained in the k+1th iteration.

[0133] During the iteration process, the increase of the inertia weight coefficient ω can improve the global search ability of the particle swarm algorithm, and the decrease of the inertia weight coefficient will enhance the algorithm's search accuracy for local extreme values. Therefore, it is more reasonable to make the inertia weight decrease linearly during the iteration process. This strategy can enable the algorithm to diffuse particles to the entire solution space more quickly in the early stage of the iteration, and to keep them concentrated in the area near the optimal solution in the later stage of the iteration, thereby improving the search accuracy for the optimal solution. When the inertia weight coefficient decreases linearly from 0.9 to 0.4, a good optimization effect can usually be obtained. Therefore, the weight coefficient is designed to be

[0134]

[0135] where ω start is the initial inertia weight, set to 0.9; ω end is the inertia weight when iterating to the maximum evolutionary generation, and the setting range is 0.4; k is the number of iterations, which is set to 30 times; T max is the maximum number of iterations.

[0136] According to the above iterative method, the particles of the particle swarm algorithm will move around so that the speed ITAE error gradually decreases. After multiple iterations, the particle swarm will tend to the global optimal solution as a whole.

[0137] The particle swarm iteration is as follows Figure 8 The blue curve is the result of optimizing only the proportional factor of the fuzzy controller, and the red curve is the result of optimizing both the proportional factor and the fuzzy rule weight. It can be seen that optimizing the fuzzy rule weight can make the fuzzy controller have a more refined control effect and a faster speed response. The optimized proportional factor is 1.048, and the fuzzy rule weight is shown in Table 2.

[0138] Table 2

[0139]

[0140] Substitute the optimized fuzzy controller parameters into the actual fuzzy controller to realize the linearization of voltage feedback gain. When the motor enters the low-speed weak magnetic field area, the adaptive controller adjusts the voltage feedback loop gain, so that the sensitivity of current angle regulation increases and the voltage regulation speed increases; under medium and high speed weak magnetic field conditions, the adaptive controller adaptively adjusts the integral coefficient to ensure the stability of current even under small voltage margin.

[0141] The adaptive weakening magnetic field control algorithm proposed in the present invention first designs an adaptive controller structure, proposes an adaptive coefficient, and decouples it into a speed adjustment factor and a current control angle adjustment factor; then an inverse function is used to construct a speed adjustment factor, and fuzzy control is used to construct a current control angle adjustment factor to eliminate the influence of speed changes and current changes on the voltage feedback gain; and a particle swarm algorithm is used to optimize the fuzzy controller to enhance the effect of the fuzzy controller, so that the motor has better dynamic response characteristics in the low-speed weakening magnetic field area; and the jitter of voltage and current can be reduced in the high-speed weakening magnetic field area.

[0142] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A permanent magnet synchronous motor control method based on adaptive field weakening control, characterized in that: The following steps are involved: 11) Design of adaptive controller: Design adaptive coefficients to transform three non-independent influencing factors into two independent influencing factors: speed and current angle; 12) Constructing the speed adjustment factor: Using an inverse function to construct the speed adjustment factor to eliminate the influence of speed change on the voltage feedback gain; 13) Construct current angle adjustment factor: Design a fuzzy controller and use fuzzy control to construct the current control angle adjustment factor to eliminate the influence of current change on voltage feedback gain; 14) Use particle swarm algorithm to optimize fuzzy control: Offline optimization of the output proportional factor and fuzzy rule weight of the fuzzy controller is performed to eliminate the randomness of the parameters of the fuzzy controller during design; The optimized fuzzy controller parameters are substituted into the actual fuzzy controller to realize the linearization of voltage feedback gain; when the motor enters the low-speed weak magnetic field area, the adaptive controller adjusts the voltage feedback loop gain, so that the sensitivity of current angle regulation increases and the voltage regulation speed is accelerated; under medium and high speed weak magnetic field conditions, the adaptive controller adaptively adjusts the integral coefficient to ensure the stability of current even under small voltage margin.

2. A permanent magnet synchronous motor control method based on adaptive field weakening control according to claim 1, characterized in that: The design of the adaptive controller includes the following steps: 21) For each steady-state operating point, the gain of the voltage feedback loop is set as follows: Among them, ω e is the motor electrical angular velocity, I s is the motor stator current amplitude, β is the motor current control angle, U s is the motor stator voltage amplitude; Since at a certain steady-state operating point, the voltage limit circle is determined as the speed is determined, when the current is determined, there must be a unique current control angle corresponding to it, so ω in the above formula is e ,I s and β are not independent in the weak magnetic region. If ω e With β and β as the two independent variables in the above equation, the gain of the voltage feedback loop is obviously nonlinear for these two independent variables; 22) In order to obtain the control bandwidth and stability margin at each static operating point, the adaptive coefficient K is proposed adapt : Based on the relatively slow change in speed and the rapid change in current, K adapt Decoupled into speed adjustment factor K ω and current control angle adjustment factor K β , the gain of the voltage feedback loop is kept almost constant by these two adjustment factors: K adapt =K ω ·K β , Among them, K adapt is the adaptive coefficient, K ω is the speed adjustment factor, K β is the current control angle adjustment factor, C is a constant; 23) Design of adaptive controller: Adaptive coefficient is decoupled into speed adjustment factor K ω and current control angle adjustment factor K β , the current regulator and speed regulator output K respectively β and K ω , adjust the gain of the voltage feedback loop, and the output current control angle passes through the anti-saturation limiter to obtain the current control angle required for weak magnetic field.

3. The method for controlling a permanent magnet synchronous motor based on adaptive field weakening control according to claim 1, characterized in that: The construction of the rotation speed adjustment factor comprises the following steps: 31) Based on the fact that the maximum speed of the built-in permanent magnet synchronous motor in actual application can reach 3 to 4 times the base speed, assuming that the maximum speed is 3 times the base speed, the gain of the voltage control loop at the maximum speed is 9 times that at the base speed; in order to amplify the gain at low speed, the speed adjustment factor K ω The design is as follows: Where n and ω are the speed and angular velocity of the motor, n b and ω b It represents the rotation speed and angular velocity of the motor at the base speed, and k represents the ratio of the maximum speed to the base speed; 32) When other conditions remain unchanged, the speed adjustment factor K ω To make the gain of the voltage control loop constant, when the motor operates below the base speed, that is, below the maximum speed at maximum torque output, the reduction in the denominator leads to K ω Increase; when the motor speed is lower than the basic speed, the limiter works so that the entire weak magnetic control loop has no effect on the motor and does not affect the stability of the control loop.

4. The method for controlling a permanent magnet synchronous motor based on adaptive field weakening control according to claim 1, characterized in that: The construction of the current angle adjustment factor comprises the following steps: 41) Fuzzy control is used to control the current angle adjustment factor. The input is the current control angle and its rate of change, and the output is the adjustment factor K. fuzzy , which describes the effect of the current control angle change on the gain of the weak magnetic feedback loop. At the same time, the current control angle adjustment factor K β satisfy: 42) Design of fuzzy controller: The input of the fuzzy controller is the current control angle and its rate of change. The input is mapped from the physical domain to the fuzzy domain through the quantization factor to obtain a clear value. The clear value of the input is converted to the corresponding fuzzy subset and membership function through fuzzification. According to the fuzzy rules, the fuzzification is resolved to obtain the exact value, and then the output variable is mapped from the fuzzy domain to the physical domain by the proportional factor, and the output is the adjustment coefficient. 43) Set the input and output of the fuzzy controller: The input of the fuzzy controller is a two-dimensional vector consisting of β and Δβ, and the output is the adjustment factor K fuzzy , Δβ is the derivative of β, β and Δβ are respectively determined by the quantization factor K i1 and K i2 Mapped to [1,2] and [-1,1], K fuzzy Finally, it is mapped to [1,10] by the scale factor; 44) Fuzzy processing of fuzzy controller: β and K fuzzy It is described by seven natural language variables, as follows: negative large NB, negative medium NM, negative small NS, zero Z, positive small PS, positive medium PM, positive large PB; Δβ is described by three linguistic variables: negative N, zero Z, positive P; 45) Fuzzy controller defuzzification: The defuzzification process uses the centroid method to find the exact value of the output variable. In the formula, x i is the value of the horizontal axis, μ N (x i ) represents the membership function at x i The value at 46) Finally, adjust the coefficient K fuzzy Substitute into 41) to calculate the current control angle adjustment factor K β .

5. The method for controlling a permanent magnet synchronous motor based on adaptive field weakening control according to claim 1, characterized in that: The optimization of the fuzzy controller by using the particle swarm algorithm comprises the following steps: 51) The particle swarm algorithm is used to optimize the fuzzy controller. In order to make the designed fuzzy controller have higher control accuracy, the proportional coefficient of fuzzy control and the weight of fuzzy rules are used as the input of the particle swarm algorithm; the initial particle position vector is defined as Where p1 i (1×1) is a one-dimensional population, K o Represents the proportional factor of the fuzzy controller; p2 i (1×21) is a 21-dimensional population, w1,w2,…,w 21 They represent the weights of fuzzy rules respectively; i represents the member number in the population, and N represents the population size, that is, the number of particles in the population; 52) The particle swarm algorithm selects the ITAE error of the speed as the objective function of the particle swarm. The effect of the control algorithm is measured by speed and torque, so the fitness function J ITAE Defined as: J ITAE =∫t·|e|·dt, In the formula, e is the speed error or torque error of the motor, and t is the action time; 53) The particle swarm algorithm generates the initial particle position of the fuzzy controller. All particles in the particle swarm adjust their speed and position for a finite number of iterations according to their individual extreme values ​​and the current global optimal solution. The specific iteration function is: In the formula, represents the d-dimensional velocity component of particle i in the k-th iteration, represents the d-dimensional position component of particle i in the k-th iteration, c1 and c2 are learning factors, which adjust the degree to which the particle draws on individual experience and group experience respectively; r1 and r2 are two random numbers that increase the randomness of the search; ω represents the weight factor for the particle to inherit its previous velocity; and are the individual optimum and global optimum obtained at the kth iteration, respectively. The individual optimum is a d-dimensional vector, and the global optimum is a value. represents the d-dimensional velocity component of particle i in the k+1th iteration, The optimal individual obtained in the k+1th iteration; 54) In the iterative process, the increase of the inertia weight coefficient ω improves the global search ability of the particle swarm algorithm, and the decrease of the inertia weight coefficient enhances the algorithm's search accuracy for local extreme values. Based on this, the weight coefficient ω(k) is designed as: Among them, ω start is the initial inertia weight, ω end is the inertia weight when iterating to the maximum evolutionary generation, k is the number of iterations, T max is the maximum number of iterations; 55) According to the iterative method of particle swarm algorithm, particles will move around so that the speed ITAE error gradually decreases. After multiple iterations, the particle swarm will tend to the global optimal solution as a whole; 56) The optimized fuzzy controller parameters are obtained by particle swarm optimization algorithm, and the obtained parameters are substituted into the actual fuzzy controller to realize voltage feedback gain linearization; When the motor enters the low-speed weak-magnetic field area, the adaptive controller adjusts the voltage feedback loop gain, which increases the sensitivity of the current angle regulation and speeds up the voltage regulation. Under medium- and high-speed weak-magnetic field conditions, the adaptive controller can adaptively adjust the integral coefficient to ensure current stability even under small voltage margins.

Images