Robust model predictive current control method for surface-mounted permanent magnet synchronous motor

Abstract

The invention provides a surface-mounted permanent magnet synchronous motor robust model prediction current control method, and the method comprises the steps: obtaining a first current prediction model from a permanent magnet synchronous motor mathematical model, and carrying out the processing of the first current prediction model, and obtaining a second current prediction model; obtaining a current prediction error according to the first and second current prediction models; inductance information and flux linkage information are extracted based on the current prediction error, a second current prediction model is corrected based on the inductance information and the flux linkage information to obtain an inductance parameter correction value and a flux linkage parameter correction value, and the voltage vector with the minimum extension cost function is combined to serve as the optimal voltage vector. Motor parameters existing in the obtained prediction model are consistent with actual parameters, good dynamic performance and steady-state performance can be kept, current harmonics and steady-state errors are effectively reduced, negative effects caused by parameter mismatch are eliminated, the parameter sensitivity of model prediction current control is reduced, and the comprehensive performance of the whole system is improved.

Description

Technical Field

[0001] The present application relates to the technical field of permanent magnet synchronous motors, and in particular to a robust model predictive current control method for a surface-mounted permanent magnet synchronous motor. Background Art

[0002] Permanent magnet synchronous motors have the advantages of high efficiency and high power density, and are widely used in high-performance electric drive applications. Among the various advanced control algorithms that have emerged to further improve control performance, permanent magnet synchronous motor finite set model predictive control has gradually become popular in recent years due to its simple algorithm and structure and good dynamic control performance.

[0003] However, finite set model predictive control is a model-based method with model parameter dependence problems. Finite set model predictive control relies heavily on the prediction model to select the optimal voltage vector, and the motor parameters in the prediction model need to be consistent with the actual parameters in order to maintain good dynamic performance and steady-state performance. When the parameters in the model do not match the actual parameters, the voltage vector selected by the cost function is not the true optimal voltage vector, which will further lead to a decrease in the control performance of the entire control system. Therefore, it is particularly critical to improve the parameter robustness of finite set model predictive control. Summary of the invention

[0004] The present application provides a surface-mounted permanent magnet synchronous motor model predictive current control method, which corrects the inductance parameter and flux linkage parameter of the permanent magnet synchronous motor in real time, and uses the inductance parameter correction value and the flux linkage parameter correction value to correct the parameters in the mathematical model of the permanent magnet synchronous motor, thereby improving the parameter robustness of the model predictive control, and includes the following steps:

[0005] S1, establish the mathematical model of permanent magnet synchronous motor.

[0006] S2, obtaining a first current prediction model from the permanent magnet synchronous motor mathematical model through the forward Euler method, performing one-beat delay compensation on the first current prediction model or adding parameter disturbance existing in the first current prediction model to obtain a second current prediction model.

[0007] S3, the second current prediction model includes the second predicted current, the first current prediction model includes the first predicted current, the second predicted current is subtracted from the first predicted current to obtain a current prediction error; based on the current prediction error, inductance information is extracted, the inductance information is brought into the second current prediction model for correction, and an inductance parameter correction value is obtained; based on the current prediction error, magnetic flux information is extracted, the magnetic flux information is brought into the second current prediction model for correction, and a magnetic flux parameter correction value is obtained.

[0008] S4, bringing the inductance parameter correction value and the flux linkage parameter correction value into the second current prediction model to correct the parameters in the permanent magnet synchronous motor mathematical model.

[0009] S5, selecting a voltage vector with the smallest extended cost function as an optimal voltage vector, converting the optimal voltage vector into a corresponding switch state to act on the inverter, and performing current control on the motor.

[0010] Specifically, the permanent magnet synchronous motor mathematical model includes:

[0011]

[0012] Among them, u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; i d 、i q are the d-axis and q-axis current components in the dq coordinate system respectively; ω e is the electrical angular velocity; R is the stator resistance; Ψ f is the permanent magnet flux amplitude; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q .

[0013] Specifically, the first current prediction model expression includes:

[0014]

[0015] Among them, i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; i d (l+1), i q (k+1) are the current components of the d-axis and q-axis at the time (k+1) in the dq coordinate system; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; f is the flux amplitude of the permanent magnet.

[0016] Specifically, the second current prediction model expression obtained by performing one-beat delay compensation on the first current model includes:

[0017]

[0018] in, are the one-beat delay compensation current components of the d-axis and q-axis in the dq coordinate system respectively; i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; f is the flux amplitude of the permanent magnet.

[0019] Specifically, the expression of the second current prediction model obtained by adding the parameter perturbation existing in the first current prediction model to the first current model includes:

[0020]

[0021] Among them, i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; i′ d (k+1), i′ q (k+1) are the current components of the d-axis and q-axis at the time (k+1) in the dq coordinate system after adding parameter disturbance; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; f is the permanent magnet flux amplitude; ΔR is the deviation of the stator resistance parameter; ΔL is the deviation of the inductance parameter; ΔΨ f is the deviation of the permanent magnet flux amplitude parameter.

[0022] Specifically, extracting inductance information based on the current prediction error, bringing the inductance information into the second current prediction model for correction, and obtaining an inductance parameter correction value, the specific steps include:

[0023] S51, the current prediction error of the d-axis is simplified to obtain:

[0024]

[0025] Among them, R0, L0, ψ f0 are the motor parameters before calibration;

[0026] S52, since the current prediction error of the d-axis only includes the inductance error, the formula in step S51 is rearranged to obtain:

[0027]

[0028] Among them, the E dL (k+1) is the d-axis current error at time k+1;

[0029] S53, define the common coefficient of the current prediction error of the d-axis as:

[0030]

[0031] Substituting the formula in step S52, the formula of the common coefficient can be transformed into:

[0032]

[0033] S54, C L (k) and E dL The transfer function G between (k+1) CL (k) can be expressed as:

[0034]

[0035] S55, by rearranging, we can get ΔL and C L The relationship between:

[0036]

[0037] Where ΔL(k) is the inductance error at time k;

[0038] S56, ΔL(k) is processed by a low-pass filter as follows:

[0039] ΔL f (k) = α·ΔL(k) + (1-α)·ΔL f (k-1),

[0040] Where, ΔLf (k) is the filter output value at time k; α is the filter coefficient, and its value is 0.02;

[0041] S57, finally, the inductance parameter correction value L0(k) is obtained:

[0042] L0(k)=L-ΔL f (k).

[0043] Specifically, the flux information is extracted based on the current prediction error, and the flux information is brought into the second current prediction model for correction to obtain a flux parameter correction value. The specific steps include:

[0044] S61, the current prediction error of the q-axis is simplified to obtain:

[0045]

[0046] Among them, R0, L0, ψ f0 are the motor parameters before calibration;

[0047] S62, since the current prediction error of the q-axis only includes the flux error, the formula in step S61 is rearranged to obtain:

[0048]

[0049] Among them, the E qψf (k+1) is the q-axis current error at time k+1;

[0050] S63, according to the formula in S62, the flux linkage error is directly extracted as:

[0051]

[0052] Among them, Δψ f (k) is the flux error at time k;

[0053] S64, Δψ f (k) is processed by a low-pass filter as follows:

[0054] Δψ fL (k) = β·Δψ f (k)+(1-β)·Δψ fL (k-1),

[0055] Among them, Δψ fL (k) is the filter output value, β is the filter coefficient;

[0056] S65, finally, the flux parameter correction value ψ is obtained f0 (k):

[0057] ψf0 (k) = ψ f -Δψ fL (k).

[0058] Specifically, the traditional cost function is reconstructed into the extended cost function, and the specific steps are as follows:

[0059] S71, introducing a unit change current error term for each cycle into the conventional cost function, wherein the unit change current error term for each cycle includes:

[0060]

[0061] Among them, K d is the current error term with unit change d; K q is the current error term with unit change of q;

[0062] S72, taking the error caused by unit change current as the coefficient β′ of the expansion term, we can get:

[0063]

[0064] Where, λ is the gain coefficient;

[0065] S73, using first-order low-pass filtering to reduce harmonic interference, obtains:

[0066] β′ d(k) =(1-γ)β′ d(k-1) +γE′ d(k)

[0067] β′ q(k) =(1-γ)β′ q(k-1) +γE′ q(k) ,

[0068] Where: γ is the filter coefficient;

[0069] S74, coefficient β′ d(k) The product of the predicted current change on the d-axis can be used to obtain the d-axis differential cost function:

[0070]

[0071] Coefficient β′ q(k) The product of the predicted current change on the q-axis can give the q-axis differential cost function:

[0072]

[0073] S75, set the threshold ε, when the current change rate is lower than this threshold, update β′ d(k) and β′ q(k),To avoid the error in the estimation of the expansion cost caused by the denominator approaching zero, the formula is as follows:

[0074] β′ d(k) =(1-γ)β′ d(k-1)

[0075] β′ q(k) =(1-γ)β′ q(k-1) ;

[0076] S76, finally, the extended cost function is obtained as:

[0077] g=(P d(k) +β d(k) ) 2 +(P q(k) +β q(k) ) 2 ,

[0078] Among them, P d(k) is the d-axis current error term, P q(k) is the q-axis current error term.

[0079] Compared with the prior art, the robust model predictive current control method for a surface-mounted permanent magnet synchronous motor of the present application corrects the inductance parameters and flux parameters of the permanent magnet synchronous motor in real time, and uses the inductance parameter correction values ​​and flux parameter correction values ​​to correct the parameters in the mathematical model of the permanent magnet synchronous motor, thereby improving the parameter robustness of the model predictive control. The motor parameters in the obtained prediction model are consistent with the actual parameters, and good dynamic performance and steady-state performance can be maintained; and the optimal voltage vector is selected in combination with the extended cost function to control the motor, effectively reducing current harmonics and steady-state errors, eliminating the negative impact of parameter mismatch, and reducing the parameter sensitivity of the model predictive current control, thereby improving the overall performance of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0081] Figure 1 It is a flow chart of a robust model predictive current control method for a surface-mounted permanent magnet synchronous motor in an embodiment of the present application;

[0082] Figure 2 It is a system control structure block diagram of a robust model predictive current control method for a surface-mounted permanent magnet synchronous motor in an embodiment of the present application;

[0083] Figure 3is a flow chart of a current information correction algorithm in an embodiment of the present application;

[0084] Figure 4 is a flow chart of a magnetic linkage information correction algorithm in an embodiment of the present application;

[0085] Figure 5 It is the phase current THD (total harmonic distortion) under the traditional MPCC (model predictive current control method) in the comparative example of this application;

[0086] Figure 6 is the phase current THD under the robust MPCC (a robust model predictive current control method for a surface-mounted permanent magnet synchronous motor) in the embodiment of the present application;

[0087] Figure 7 is the current waveform under the traditional MPCC in the comparative example of the present application;

[0088] Figure 8 is the current waveform under the robust MPCC in the embodiment of the present application;

[0089] Fig. 9 It is a comparison chart of dynamic response time under different model parameters under the traditional MPCC and robust MPCC methods in the embodiments of the present application;

[0090] Fig.10 This is a THD comparison chart under different model parameters of the traditional MPCC and robust MPCC methods in the embodiments of the present application. DETAILED DESCRIPTION

[0091] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0092] The following describes the implementation methods of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, in the absence of conflict, the following embodiments and the features in the embodiments can be combined with each other. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work belong to the scope of protection of the present application.

[0093] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present application, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number and aspect described herein can be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this device and / or practice this method.

[0094] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application. The drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0095] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the examples can be practiced without these specific details.

[0096] like Figure 1 as well as Figure 2 As shown, this embodiment provides a robust model predictive current control method for a surface-mounted permanent magnet synchronous motor, comprising the following steps:

[0097] S1, establish the mathematical model of permanent magnet synchronous motor.

[0098] S2, obtaining a first current prediction model from the permanent magnet synchronous motor mathematical model through the forward Euler method, performing one-beat delay compensation on the first current prediction model or adding parameter disturbance existing in the first current prediction model to obtain a second current prediction model.

[0099] S3, the second current prediction model includes the second predicted current, the first current prediction model includes the first predicted current, the second predicted current is subtracted from the first predicted current to obtain a current prediction error; based on the current prediction error, inductance information is extracted, the inductance information is brought into the second current prediction model for correction, and an inductance parameter correction value is obtained; based on the current prediction error, magnetic flux information is extracted, the magnetic flux information is brought into the second current prediction model for correction, and a magnetic flux parameter correction value is obtained.

[0100] S4, bringing the inductance parameter correction value and the flux linkage parameter correction value into the second current prediction model to correct the parameters in the permanent magnet synchronous motor mathematical model.

[0101] S5, selecting a voltage vector with the smallest extended cost function as an optimal voltage vector, converting the optimal voltage vector into a corresponding switch state to act on the inverter, and performing current control on the motor.

[0102] The model structure of this embodiment is simple. It uses the current prediction error formula to directly extract inductance information and flux information, corrects the prediction model, and combines it with the extended cost function to effectively reduce current harmonics and steady-state errors, eliminate the negative impact of parameter mismatch, and improve the overall performance of the entire system.

[0103] In one embodiment, according to the mathematical model of the surface-mounted permanent magnet synchronous motor, the voltage equation in the dq coordinate system can be obtained:

[0104]

[0105] Among them, u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; i d 、i q are the d-axis and q-axis current components in the dq coordinate system respectively; ω e is the electrical angular velocity; R is the stator resistance; Ψ f is the permanent magnet flux amplitude; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q .

[0106] The torque equation is: Where p is the number of motor pole pairs, Ψ f is the flux amplitude of the permanent magnet.

[0107] In one embodiment, the first current prediction model is obtained from the permanent magnet synchronous motor mathematical model by the forward Euler method, and the voltage equation in the dq coordinate system is discretized to obtain the current i at the next moment. d (k), i q (k), so that the first current prediction model expression includes:

[0108]

[0109] Among them, i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; i d (k+1), i q(k+1) are the current components of the d-axis and q-axis at the time (k+1) in the dq coordinate system; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; f is the flux amplitude of the permanent magnet.

[0110] In one embodiment, in order to predict and correct the influence of delay and improve control accuracy, one-beat delay compensation is applied to the first current model, and the second current prediction model expression obtained by performing one-beat delay compensation on the first current model includes:

[0111]

[0112] in, are the one-beat delay compensation current components of the d-axis and q-axis in the dq coordinate system respectively; i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; f is the flux amplitude of the permanent magnet.

[0113] In one embodiment, considering the parameter disturbance existing in the model, the expression of the second current prediction model obtained by adding the parameter disturbance existing in the first current prediction model to the first current model includes:

[0114]

[0115] Among them, i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; i′ d (k+1), i′ q(k+1) are the current components of the d-axis and q-axis at the time (k+1) in the dq coordinate system after adding parameter disturbance; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; f is the permanent magnet flux amplitude; ΔR is the deviation of the stator resistance parameter; ΔL is the deviation of the inductance parameter; ΔΨ f is the deviation of the permanent magnet flux amplitude parameter.

[0116] like Figure 3 As shown, in one embodiment, the current prediction error includes a d-axis current prediction error and a q-axis current prediction error. Based on the d-axis current prediction error in the current prediction error, inductance information can be extracted, and the inductance information is brought into the second current prediction model for correction to obtain an inductance parameter correction value. The specific steps include:

[0117] S51, the value of TR / L in the d-axis current prediction error can be ignored, so the d-axis current prediction error is simplified to:

[0118]

[0119] Among them, R0, L0, ψ f0 are the motor parameters before calibration;

[0120] S52, since the current prediction error of the d-axis only includes the inductance error, the formula in step S51 is rearranged to obtain:

[0121]

[0122] Among them, the E dL (k+1) is the d-axis current error at time k+1;

[0123] S53, define the common coefficient of the current prediction error of the d-axis as:

[0124]

[0125] Substituting the formula in step S52, the formula of the common coefficient can be transformed into:

[0126]

[0127] S54, C L (k) and E dL The transfer function G between (k+1) CL (k) can be expressed as:

[0128]

[0129] S55, by rearranging, we can get ΔL and C L The relationship between:

[0130]

[0131] Where ΔL(k) is the inductance error at time k;

[0132] S56, to reduce harmonic interference, ΔL(k) is processed by a low-pass filter as follows:

[0133] ΔL f (k) = α·ΔL(k) + (1-α)·ΔL f (k-1),

[0134] Among them, ΔL f (k) is the filter output value at time k; α is the filter coefficient, which can be 0.02;

[0135] S57, finally, the inductance parameter correction value L0(k) is obtained:

[0136] L0(k)=L-ΔL f (k).

[0137] like Figure 5 As shown, in one embodiment, the current prediction error includes a d-axis current prediction error and a q-axis current prediction error, and the flux information is extracted based on the q-axis current prediction error in the current prediction error, and the flux information is brought into the second current prediction model for correction to obtain a flux parameter correction value. The specific steps include:

[0138] S61, the current prediction error of the q-axis is simplified to obtain:

[0139]

[0140] Among them, R0, L0, ψ f0 are the motor parameters before calibration;

[0141] S62, since the current prediction error of the q-axis only includes the flux error, the formula in step S61 is rearranged to obtain:

[0142]

[0143] Among them, the E qψf (k+1) is the q-axis current error at time k+1;

[0144] S63, according to the formula in S62, the flux linkage error is directly extracted as:

[0145]

[0146] Among them, Δψ f (k) is the flux error at time k;

[0147] S64, in order to reduce harmonic interference, Δψ f (k) is processed by a low-pass filter as follows:

[0148] Δψ fL (k) = β·Δψ f (k)+(1-β)·Δψ fL (k-1),

[0149] Among them, Δψ fL (k) is the filter output value, β is the filter coefficient;

[0150] S65, finally, the flux parameter correction value ψ is obtained f0 (k):

[0151] ψ f0 (k) = ψ f -Δψ fL (k).

[0152] In one embodiment, after directly extracting the motor information, the corrected motor information is compensated into the prediction model, which can theoretically eliminate the steady-state error. However, due to the discrete characteristics of the current control switch predicted by the permanent magnet synchronous motor model, each beat predicted by the model may produce a control error, so the traditional cost function is reconstructed into the extended cost function, and the specific steps are as follows:

[0153] S71, in order to compensate for the control error of each beat, a unit change current error term of each cycle is introduced into the traditional cost function, and the unit change current error term of each cycle includes:

[0154]

[0155] Among them, K d is the current error term with unit change; K q is the current error term with unit change of q;

[0156] S72, taking the error caused by unit change current as the coefficient β′ of the expansion term, we can get:

[0157]

[0158] Where, λ is the gain coefficient;

[0159] S73, in order to reduce harmonic interference, adopt first-order low-pass filtering to reduce harmonic interference:

[0160] β′ d(k) =(1-γ)β′ d(k-1) +γE′ d(k)

[0161] β′ q(k) =(1-γ)β′ q(k-1) +γE′ q(k) ,

[0162] Where: γ is the filter coefficient;

[0163] S74, coefficient β′ d(k) The product of the predicted current change on the d-axis can be used to obtain the d-axis differential cost function:

[0164]

[0165] Coefficient β′ q(k) The product of the predicted current change on the q-axis can give the q-axis differential cost function:

[0166]

[0167] S75, to avoid the estimation error of the extended cost function caused by the denominator approaching zero, set the threshold ε, and update β′ when the current change rate is lower than this threshold. d(k) and β′ q(k) ,To avoid the error in the estimation of the expansion cost caused by the denominator approaching zero, the formula is as follows:

[0168] β′ d(k) =(1-γ)β′ d(k-1)

[0169] β′ q(k) =(1-γ)β′ q(k-1) ;

[0170] S76, finally, the extended cost function is obtained as:

[0171] g=(P q(k) +β d(k) ) 2 +(P q(k) +β q(k) ) 2 ,

[0172] Among them, P d(k) is the d-axis current error term, P q(k) is the q-axis current error term.

[0173] Example 1

[0174] The robust model predictive current control method for a surface-mounted permanent magnet synchronous motor provided in the above embodiment is simulated and modeled in the Matlab / Simulink environment. The specific steps are as follows:

[0175] S1, establish the mathematical model of permanent magnet synchronous motor The parameters of the permanent magnet synchronous motor (prediction model) are set as follows: DC bus voltage is 311V, pole pair number P is 4, stator resistance is 1.5Ω, stator inductance is 11.1mH, and rotor flux is 0.27Wb. The rated speed of the permanent magnet synchronous motor is set to 750r / min and the rated torque is set to 9N·m.

[0176] S2, obtaining a first current prediction model from the permanent magnet synchronous motor mathematical model by using a forward Euler method Perform one-beat delay compensation on the first current prediction model or add parameter disturbance existing in the first current prediction model to obtain a second current prediction model

[0177] S3, the second current prediction model includes a second predicted current, the first current prediction model includes a first predicted current, and the second predicted current is subtracted from the first predicted current to obtain a current prediction error

[0178] Extracting inductance information based on the current prediction error The inductance information is brought into the second current prediction model for correction to obtain the inductance parameter correction value L0(k)=L-ΔL f (k); Extracting flux information based on the current prediction error The flux information is brought into the second current prediction model for correction to obtain a flux parameter correction value ψ f0 (k) = ψ f -Δψ fL (k);

[0179] S4, the inductance parameter correction value L0(k)=L-ΔL f (k) and the flux parameter correction value ψ f0 (k) = ψ f -Δψ fL (k) Introduce the second current prediction model Correct the parameters in the mathematical model of permanent magnet synchronous motor;

[0180] S5, select the extended cost function The minimum voltage vector is used as the optimal voltage vector, and the optimal voltage vector is converted into a corresponding switching state to act on the inverter to control the current of the motor.

[0181] like Figure 6 As shown, it can be seen that the total harmonic distortion of phase current under a robust model prediction current control method of a surface-mounted permanent magnet synchronous motor, and the phase current THD (total harmonic distortion) of the robust MPCC method is 9.05%.

[0182] like Figure 8 As shown in the figure, the current waveform under the robust model predictive current control method (robust MPCC) of a surface-mounted permanent magnet synchronous motor can be seen. When the parameters are matched, the robust MPCC has good control performance. Under the robust MPCC control method, the d-axis and q-axis current prediction errors are only 0.32A and 0.22A.

[0183] To further verify the effectiveness of the robust MPCC method proposed in this invention, the motor speed is set to 750 r / min, the torque is reduced from 9 N·m to 6 N·m, and the resistance, flux linkage and inductance parameters in the prediction model are suddenly changed to 0.5 times and 2 times of the original ones, respectively, to simulate the dynamic performance and steady-state performance of the robust MPCC method under harsh parameters.

[0184] like Fig. 9 As shown, the dynamic performance under the robust MPCC method can be seen. The dynamic response time is 35, 36, 37, 35, 34, 32, and 35 respectively under the same parameter conditions as in Control Example 1.

[0185] like Fig.10 As shown, the THD (total harmonic distortion) situation under the robust MPCC method can be seen. Under the same parameter conditions in Control Example 1, the THD (total harmonic distortion) situations are 9.06%, 9.02%, 9.21%, 10.39%, 9.18%, 8.88% and 9.99% respectively.

[0186] Comparative Example 1

[0187] The steps of traditional MPCC are as follows: establish a permanent magnet synchronous motor model under dq axis coordinates, measure the current current and rotor position, obtain the dq axis current through coordinate transformation, predict all possible current values ​​at the next moment based on the reference current and current state, correspond to different voltage vectors, calculate the cost function corresponding to each voltage vector, select the voltage vector that minimizes the cost function, and generate the corresponding PWM (pulse width modulation) signal, apply the signal to the selected voltage vector to the inverter, thereby driving the motor.

[0188] The parameters of the permanent magnet synchronous motor (prediction model) are set as follows: DC bus voltage is 311V, pole pair number P is 4, stator resistance is 1.5Ω, stator inductance is 11.1mH, and rotor flux is 0.27Wb. The rated speed of the permanent magnet synchronous motor is set to 750r / min and the rated torque is 9N·m. The traditional MPCC method is simulated and modeled in the Matlab / Simulink environment.

[0189] like Figure 5 As shown, the total harmonic distortion of the phase current under the traditional MPCC method can be seen. The total harmonic distortion of the phase current under the traditional MPCC method is that the phase current THD of the traditional MPCC method is 12.57%.

[0190] like Figure 7 As shown in the figure, the current waveform under the traditional MPCC test can be seen. Under the traditional MPCC control method, the d-axis and q-axis current prediction errors are as high as 1.67A and 1.89A.

[0191] The motor speed is set to r / min, the torque is reduced from 9N·m to 6N·m, and the resistance, flux linkage and inductance parameters in the prediction model are mutated to 0.5 times and 2 times of the original values ​​respectively to simulate the harsh parameters and obtain the dynamic performance and steady-state performance of the traditional MPCC method.

[0192] like Fig. 9 As shown, the dynamic performance of the traditional MPCC method can be seen. The dynamic response time is 39, 41, 45, 59, 39, 35, and 36 under different parameter conditions.

[0193] like Fig.10 As shown, the THD (total harmonic distortion) situation under the traditional MPCC method can be seen. The THD (total harmonic distortion) situations under different parameter conditions are 12.57%, 13.01%, 12.76%, 17.76%, 13.04%, 12.86%, and 12.06% respectively.

[0194] In summary, a comparative analysis is performed based on the simulation results of Control Example 1 and Embodiment 1.

[0195] Figure 5 and Figure 6 By comparison, it can be seen that when the parameters are matched, both the control example 1 and the embodiment 1 have good control performance. The phase current THD of the control example 1 is 12.57%, and the phase current THD of the embodiment 1 is 9.05%. The phase current THD of the embodiment 1 is reduced by 28% compared with the phase current THD of the control example 1, indicating that the robust MPCC method of the embodiment 1 can improve the current waveform and improve the steady-state performance of the system.

[0196] Figure 7 and Figure 8 By comparison, the d-axis and q-axis current prediction errors in Control Example 1 are as high as 1.67A and 1.89A, while the d-axis and q-axis current prediction errors in Example 1 are only 0.32A and 0.22A, indicating that the robust MPCC in Example 1 can reduce the predicted current error and improve the accuracy of the predicted current.

[0197] like Fig. 9 As shown, under harsh parameters, the dynamic response time of the traditional MPCC method in Control Example 1 is 39, 41, 45, 59, 39, 35, and 36 under different parameter conditions. The dynamic response time of the robust MPCC method in Example 1 is 35, 36, 37, 35, 34, 32, and 35 under the same parameter conditions in Control Example 1. The dynamic performance of the robust MPCC method in Example 1 is significantly better than that of the traditional MPCC method in Control Example 1 under any parameters.

[0198] like Fig.10 As shown, the THD conditions under the traditional MPCC method in Control Example 1 are 12.57%, 13.01%, 12.76%, 17.76%, 13.04%, 12.86%, and 12.06% respectively under different parameter conditions. The THD conditions under the robust MPCC method in Example 1 are 9.06%, 9.02%, 9.21%, 10.39%, 9.18%, 8.88%, and 9.99% respectively under the same parameter conditions in Control Example 1. Due to the real-time correction of the system inductance and flux parameters by the robust MPCC in Example 1, the entire system can still maintain good control performance when the parameters are mismatched. In addition, when the inductance parameter is 0.5L0, the dynamic performance and steady-state performance of the traditional MPCC method in Control Example 1 are particularly poor.

[0199] The robust MPCC in Example 1 provided in the present application has good comprehensive performance when the motor parameters do not match, which verifies the effectiveness of the robust model predictive current control method for a surface-mounted permanent magnet synchronous motor provided in this embodiment.

[0200] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A robust model predictive current control method for a surface mounted permanent magnet synchronous motor, characterized in that: include: S1, establish a mathematical model of permanent magnet synchronous motor; S2, obtaining a first current prediction model from the permanent magnet synchronous motor mathematical model by using the forward Euler method, performing one-beat delay compensation on the first current prediction model or adding parameter disturbance existing in the first current prediction model to obtain a second current prediction model; S3, the second current prediction model includes a second predicted current, the first current prediction model includes a first predicted current, and the second predicted current is subtracted from the first predicted current to obtain a current prediction error; Extracting inductance information based on the current prediction error, bringing the inductance information into the second current prediction model for correction, and obtaining an inductance parameter correction value; Extracting flux information based on the current prediction error, bringing the flux information into the second current prediction model for correction, and obtaining a flux parameter correction value; S4, bringing the inductance parameter correction value and the flux linkage parameter correction value into the second current prediction model to correct the parameters in the permanent magnet synchronous motor mathematical model; S5, selecting a voltage vector with the smallest extended cost function as an optimal voltage vector, converting the optimal voltage vector into a corresponding switch state to act on the inverter, and performing current control on the motor.

2. The robust model predictive current control method for a surface mounted permanent magnet synchronous motor according to claim 1, characterized in that: The permanent magnet synchronous motor mathematical model includes: Among them, u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; i d 、i q are the d-axis and q-axis current components in the dq coordinate system respectively; ω e is the electrical angular velocity; R is the stator resistance; Ψ f is the permanent magnet flux amplitude; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q .

3. The robust model predictive current control method for a surface mounted permanent magnet synchronous motor according to claim 1, characterized in that: The first current prediction model expression includes: Among them, i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; i d (k+1), i q (k+1) are the current components of the d-axis and q-axis at the time (k+1) in the dq coordinate system; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; Ψ f is the flux amplitude of the permanent magnet.

4. The robust model predictive current control method for a surface mounted permanent magnet synchronous motor according to claim 1, characterized in that: The second current prediction model expression obtained by performing one-beat delay compensation on the first current model includes: in, are the one-beat delay compensation current components of the d-axis and q-axis in the dq coordinate system respectively; i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; Ψ f is the flux amplitude of the permanent magnet.

5. The robust model predictive current control method for a surface mounted permanent magnet synchronous motor according to claim 1, characterized in that: The expression of the second current prediction model obtained by adding the parameter perturbation existing in the first current prediction model to the first current model includes: Among them, i d (k), i q (k) are the current components of the d-axis and q-axis at time k in the dq coordinate system; i′ d (k+1), i′ q (k+1) are the current components of the d-axis and q-axis at the time (k+1) in the dq coordinate system after adding parameter disturbance; T is the control period; R is the stator resistance; L d , L q They are the d-axis and q-axis inductance components in the dq coordinate system, respectively. The d-axis inductance component and the q-axis inductance component are equal: L = L d =L q ;ω e is the electrical angular velocity; u d 、u q are the d-axis and q-axis voltage components in the dq coordinate system respectively; Ψ f is the permanent magnet flux amplitude; ΔR is the deviation of the stator resistance parameter; ΔL is the deviation of the inductance parameter; ΔΨ f is the deviation of the permanent magnet flux amplitude parameter.

6. The robust model predictive current control method for a surface mounted permanent magnet synchronous motor according to claim 5, characterized in that: Extracting inductance information based on the current prediction error, bringing the inductance information into the second current prediction model for correction, and obtaining an inductance parameter correction value, the specific steps include: S51, the current prediction error of the d-axis is simplified to obtain: Among them, R0, L0, ψ f0 are the motor parameters before calibration; S52, since the current prediction error of the d-axis only includes the inductance error, the formula in step S51 is rearranged to obtain: Among them, the E dL (k+1) is the d-axis current error at time k+1; S53, define the common coefficient of the current prediction error of the d-axis as: Substituting the formula in step S52, the formula of the common coefficient can be transformed into: S54, C L (k) and E dL The transfer function G between (k+1) CL (k) can be expressed as: S55, by rearranging, we can get ΔL and C L The relationship between: Where ΔL(k) is the inductance error at time k; S56, ΔL(k) is processed by a low-pass filter as follows: ΔL f (k)=α·ΔL(k)+(1-α)·ΔL f (k-1), Among them, ΔL f (k) is the filter output value at time k; α is the filter coefficient, and its value is 0.02; S57, finally, the inductance parameter correction value L0(k) is obtained: L0(k)=L-ΔL f (k) 7. The robust model predictive current control method for a surface mounted permanent magnet synchronous motor according to claim 5, characterized in that: Extracting flux information based on the current prediction error, bringing the flux information into the second current prediction model for correction, and obtaining a flux parameter correction value, the specific steps include: S61, the current prediction error of the q-axis is simplified to obtain: Among them, R0, L0, ψ f0 are the motor parameters before calibration; S62, since the current prediction error of the q-axis only includes the flux error, the formula in step S61 is rearranged to obtain: Among them, the E qψf (k+1) is the q-axis current error at time k+1; S63, according to the formula in S62, the flux linkage error is directly extracted as: Among them, Δψ f (k) is the flux error at time k; S64, Δψ f (k) is processed by a low-pass filter as follows: △ψ fL (k)=β·△ψ f (k)+(1-β)·△ψ fL (k-1), Among them, Δψ fL (k) is the filter output value, β is the filter coefficient; S65, finally, the flux parameter correction value ψ is obtained f0 (k): ψ f0 (k)=ψ f -△ψ fL (k)。 8. The robust model predictive current control method for a surface mounted permanent magnet synchronous motor according to claim 1, characterized in that: The traditional cost function is reconstructed into the extended cost function, and the specific steps are as follows: S71, introducing a unit change current error term for each cycle into the conventional cost function, wherein the unit change current error term for each cycle includes: Among them, K d is the current error term with unit change d; K q is the current error term with unit change of q; S72, taking the error caused by unit change current as the coefficient β′ of the expansion term, we can get: Where, λ is the gain coefficient; S73, using first-order low-pass filtering to reduce harmonic interference, obtains: b′ d(k) =(1-γ)β′ d(k-1) +γE′ d(k) b′ q(k) =(1-γ)β′ q(k-1) +γE′ q(k) , Where: γ is the filter coefficient; S74, coefficient β′ d(k) The product of the predicted current change on the d-axis can be used to obtain the d-axis differential cost function: Coefficient β′ q(k) The product of the predicted current change on the q-axis can give the q-axis differential cost function: S75, set the threshold ε, when the current change rate is lower than this threshold, update β′ d(k) and β′ q(k) ,To avoid the error in the estimation of the expansion cost caused by the denominator approaching zero, the formula is as follows: b′ d(k) =(1-γ)β′ d(k-1) b′ q(k) =(1-γ)β′ q(k-1) ; S76, finally, the extended cost function is obtained as: g=(P d(k) +b d(k) ) 2 +(P q(k) +b q(k) ) 2 , Among them, P d(k) is the d-axis current error term, P q(k) is the q-axis current error term.

Images