Model prediction control strategy based on voltage source type three-phase inverter

Abstract

The invention relates to a model prediction control strategy based on a voltage source type three-phase inverter, and the control strategy specifically comprises the following steps: S1, building a three-phase voltage source grid-connected inverter model, thereby obtaining a voltage vector; s2, establishing a load model so as to determine a current vector under a three-phase balanced load and a counter electromotive force of the load; s3, constructing a discrete time domain model, and predicting the load current value of the next sampling period by using the current and voltage measurement values of the current sampling period; and S4, outputting the most appropriate voltage Vabc so as to realize predictive control. The control strategy can improve the control performance of the three-phase inverter.

Description

Technical Field

[0001] The present invention relates to the technical field of inverters, and more particularly to a model predictive control strategy based on a voltage source type three-phase inverter. Background Art

[0002] The traditional control schemes of three-phase inverters mainly include hysteresis current control schemes and pulse width modulation (PWM) control schemes. The former uses the measured current and current reference value for hysteresis comparison to generate control signals, so the control scheme is simple in concept and has good dynamic performance. However, the phases are not completely decoupled, and the switching frequency will change according to the changes in load parameters and operating conditions, which is easy to cause resonance problems and also limits its application in high-power inverters. The latter generates a load voltage reference value by using the proportional-integral (PI) regulator based on the difference between the output current reference value and the actual current value, and sends it to the PWM generation module to generate control pulses. Therefore, the switching frequency is fixed, and the control performance depends on the controller parameter design. At the same time, the input sinusoidal reference error is easy to increase with the increase of the reference current frequency, which may be unacceptable in some applications. Therefore, the control performance of the three-phase inverter is not high.

[0003] The patent publication number is CN201711005038.8, which discloses a model predictive control strategy for a discrete inductive parallel interleaved inverter. The optimal output level of the phase voltage is determined by taking the phase voltage output level as the value function of the control element. For the control of each phase in the discrete inductive PII, only a maximum of 3 rolling optimization calculations are required in each sampling period, and the amount of calculation is greatly reduced. The control strategy can be expanded to any number of output levels without being restricted by the number of parallel bridge arms N, and adapts to the advantage of flexible modulation of the discrete inductive PII structure.

[0004] The technical solutions disclosed in the prior art are mainly to solve the problems of system safety and stability, but cannot improve the control performance. Summary of the invention

[0005] The technical problem solved by the present invention is to overcome the problems of the prior art and provide a model predictive control strategy based on a voltage source type three-phase inverter with high control performance.

[0006] The purpose of the present invention is achieved through the following technical solutions:

[0007] A model predictive control strategy based on a voltage source three-phase inverter is disclosed, and the control strategy specifically includes the following steps:

[0008] S1: Establish a three-phase voltage source grid-connected inverter model to obtain the voltage vector;

[0009] S2: Establish a load model to determine the current vector and the back electromotive force of the load under a three-phase balanced load;

[0010] S3: Build a discrete time domain model to use the current and voltage measurements of the current sampling period to predict the load current value of the next sampling period;

[0011] S4: Output the most suitable voltage U abc , thus achieving predictive control.

[0012] Preferably, the step S1 comprises the following steps:

[0013] S11: Establishing an inverter switch state model;

[0014] S12: Establish a voltage vector model.

[0015] Preferably, the switching state of the inverter is determined by the drive signal S a , S b and S c Decide:

[0016]

[0017] It is expressed in vector form as follows:

[0018]

[0019] In the above formula,

[0020]

[0021] The output voltage space vector u of the inverter is:

[0022]

[0023] Where: u aN ,u bN and u cN is the neutral point voltage of the inverter;

[0024] The relationship between vector u and S is:

[0025] u=U dc S

[0026] Where U dc is the DC voltage;

[0027] Considering the driving signal S a , S b and S c A total of 8 switching states are formed, and 8 voltage vectors are obtained from them, among which the zero vector U0=U7=0, so there are 7 effective voltage vectors.

[0028] Preferably, the step S2 specifically includes the following steps:

[0029] S21: Establishing a current vector model;

[0030] S22: Establishing back electromotive force model;

[0031] S23: Establish a load current dynamic model.

[0032] Preferably, in step S21, the current vector under three-phase balanced load is defined as:

[0033]

[0034] In the formula, i a 、i b 、i c They are the currents of phase a load, phase b load and phase c load respectively.

[0035] Preferably, in step S22, the back electromotive force of the load can be defined as:

[0036]

[0037] Where: R and L are load impedance and filter inductance; T s is the sampling time of the prediction algorithm; x old is the optimal vector at time k-1; i old is the measured current at time k-1.

[0038] Preferably, in step S23, the load current dynamics is described by the following vector equation:

[0039]

[0040] Where u is the inverter output voltage vector.

[0041] Preferably, the step S3 specifically includes the following steps:

[0042] S31: Establish load current differential model;

[0043] S32: current measurement value at sampling instant k;

[0044] S33: current reference at sampling instant k;

[0045] S34: establishing an iterative current model;

[0046] S35: establishing a voltage vector model;

[0047] S36: Establishing a future load current model;

[0048] S37: Cost function model.

[0049] Preferably, based on the sampling period T s After the load current vector dynamic vector equation is discretized, the current and voltage measurement values ​​of the current sampling period can be used to predict the load current value of the next sampling period;

[0050] The load current differential can be discretized as follows:

[0051]

[0052] Where i k is the measured current of the current sampling period;

[0053] The formula for the current measurement value at the sampling instant k is as follows:

[0054] i k =I meas(1) +j*I meas(2)

[0055] In the formula, I meas(1) is the transverse component of the measured current, I meas(2) To measure the longitudinal component of the current;

[0056] The current reference value formula at the sampling moment k is as follows:

[0057] i kref =I ref(1) +j*I ref(2)

[0058] In the formula, I ref(1) is the transverse component of the current reference value, I ref(2) is the longitudinal component of the current reference value;

[0059] The iterative current model stores the measured current for the next iteration:

[0060] i old =i k

[0061] The voltage vector model is used for the i-th voltage vector of current prediction:

[0062] u o1 =u(i)

[0063] In the formula, the value of i ranges from 1 to 8;

[0064] The current forecast value at time k+1 is calculated using the future load current model:

[0065]

[0066] In the MPC algorithm, for 7 different effective voltage vectors, 7 different current prediction values ​​are given; and the corresponding voltage vector whose current prediction value is closest to the current reference value will be implemented at the next sampling moment; the basic requirement of the evaluation function is to minimize the absolute value error between the reference current and the next moment sampling current; the selected effective voltage vector will be the voltage vector that minimizes the cost function;

[0067] The cost function calculation formula is as follows:

[0068] g=|real(i kref -i k1 )|+|imag(i kref -i k1 )|

[0069] In the formula, real means taking the real part of the vector, imag means taking the imaginary part of the vector; abs means taking the absolute value.

[0070] Preferably, the step S4 specifically includes the following steps:

[0071] S41: Sample the load current I generated by the voltage source inverter LOAD , the pre-selected reference current peak value I refpeak and reference frequency f ref and phase angle δ are input into the three-phase sinusoidal generator to generate the reference current I ref ;

[0072] S42:I ref and I LOAD After Clark transformation to the two-phase stationary αβ coordinate system, the reference current I ref and the measured current I meas ;

[0073] S43: The reference current I ref and the measured current I meas Bring it into the current inner loop MPC control unit and get 6 output switch states S a , S b , S c , S d , S e , S f ;

[0074] S44: According to the obtained 6 output switch states S a , S b , S c , S d , S e , S f Input to the three-phase voltage source inverter to output the most suitable voltage U abc ;

[0075] In step S43, the current inner loop MPC control process is:

[0076] S431: Obtaining the reference current value i at sampling instant k kref , which comes from the external control loop;

[0077] S432: Read the load current measurement value i at sampling instant k k ;

[0078] S433: Calculate the load back electromotive force vector e, which is used to calculate the discrete time domain model of the three-phase inverter system;

[0079] S434: Storing the measured current i for the next iteration old ;

[0080] S435: Predict the load current value i at the next sampling period, i.e., time k+1, corresponding to each different voltage vector based on the system model k1 ;

[0081] S436: Evaluate the error g between the reference current value and the predicted current value in the next sampling period based on the cost function; select the switch quantity corresponding to a vector with the smallest cost function as the input quantity of the inverter; obtain 6 output switch states S according to the optimal vector a , S b , S c , S d , S e , S f ;

[0082] S437: Outputting a voltage vector that minimizes the current error according to the six switch states as an optimal vector.

[0083] It has the following beneficial effects:

[0084] 1) Based on the mathematical model of the inverter and the load, the variables can be predicted after discretization. The switch state is substituted into the designed cost function, and the optimal drive pulse can be directly output without a modulator. Compared with the traditional control strategy, this control strategy fully utilizes the discrete characteristics of the inverter, has better current control dynamic performance, and is easy to implement. It shows better dynamic and static performance and improves control performance.

[0085] 2) Solve the problem of inappropriate output voltage vector due to large errors between current values, improve the intrinsic characteristics of the system's adaptability and robustness, and provide a new idea for the subsequent inverter control strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1A topological diagram of a grid-connected inverter based on a model predictive control strategy of a voltage source three-phase inverter according to the present invention;

[0087] Figure 2 A schematic diagram of an effective voltage vector based on a voltage source three-phase inverter of the present invention;

[0088] Figure 3 It is an algorithm flow chart of a model predictive control strategy based on a voltage source three-phase inverter of the present invention;

[0089] Figure 4 An output state diagram of a switch tube based on a model predictive control strategy of a voltage source three-phase inverter according to the present invention;

[0090] Figure 5 This is a waveform diagram of the output load A-phase voltage based on a model predictive control strategy of a voltage source three-phase inverter according to the present invention;

[0091] Figure 6 It is a waveform diagram of the output load A-phase current based on a model predictive control strategy of a voltage source three-phase inverter of the present invention;

[0092] Figure 7 The present invention is a flow chart of a model predictive control strategy based on a voltage source three-phase inverter. DETAILED DESCRIPTION

[0093] In order to clearly illustrate the technical features of the present solution, the present invention is described in detail below through specific implementation methods and in conjunction with the accompanying drawings.

[0094] In the following description, many specific details are set forth to facilitate a full understanding of the present application. However, the present application may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present application is not limited to the specific embodiments disclosed below.

[0095] In addition, in the description of the present application, it should be understood that the terms "center", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "axial", "radial", "circumferential" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first" and "second" are used only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0096] In this application, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection, an electrical connection, or a communication; it can be a direct connection, or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0097] In the present application, unless otherwise clearly specified and limited, a first feature "above" or "below" a second feature may be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in an appropriate manner in any one or more embodiments or examples.

[0098] Example 1

[0099] A model predictive control strategy based on a voltage source three-phase inverter is disclosed, and the control strategy specifically includes the following steps:

[0100] S1: Establish a three-phase voltage source grid-connected inverter model to obtain the voltage vector;

[0101] S2: Establish a load model to determine the current vector and the back electromotive force of the load under a three-phase balanced load;

[0102] S3: Build a discrete time domain model to use the current and voltage measurements of the current sampling period to predict the load current value of the next sampling period;

[0103] S4: Output the most suitable voltage U abc , thus achieving predictive control.

[0104] In this embodiment, the topology diagram of the three-phase voltage source grid-connected inverter is as follows: Figure 1 As shown in the figure, it consists of a DC power supply, an inverter, a filter inductor, a load impedance and a grid-side power supply.

[0105] Step S1 includes the following steps:

[0106] S11: Establishing an inverter switch state model;

[0107] S12: Build voltage vector model

[0108] The switching state of the inverter is determined by the drive signal S a , S b and S c Decide:

[0109]

[0110] It can be expressed in vector form as follows:

[0111]

[0112] In the above formula,

[0113] Further, the output voltage space vector u of the inverter is:

[0114]

[0115] Where: u aN ,u bN and u cN is the neutral point voltage of the inverter.

[0116] The relationship between vector u and S is:

[0117] u=U dc S

[0118] Where U dc is a DC voltage.

[0119] Considering the driving signal S a , S b and S cA total of 8 switching states are formed, and 8 voltage vectors are obtained from them, among which the zero vector U0=U7=0, so there are 7 effective voltage vectors.

[0120] In this embodiment, step S2 specifically includes the following steps:

[0121] S21: Establishing a current vector model;

[0122] S22: Establishing back electromotive force model;

[0123] S23: Establish a load current dynamic model.

[0124] The current vector under three-phase balanced load can be defined as:

[0125]

[0126] In the formula, i a 、i b 、i c They are the currents of phase a load, phase b load and phase c load respectively.

[0127] The back EMF of the load can be defined as:

[0128]

[0129] Where: R and L are load impedance and filter inductance; T s is the sampling time of the prediction algorithm; x old is the optimal vector at time k-1; i old is the measured current at time k-1.

[0130] The load current dynamics can be described by the following vector equation:

[0131]

[0132] Where u is the inverter output voltage vector.

[0133] In this embodiment, the optimal switching state is achieved and the current error is reduced, the problem of inappropriate output voltage vector due to large error between current values ​​is solved, the intrinsic characteristics of the system's adaptability and robustness are improved, and a new idea is provided for the subsequent inverter control strategy.

[0134] Example 2

[0135] The difference between this embodiment and embodiment 1 is that step S3 specifically includes the following steps:

[0136] S31: Establish load current differential model;

[0137] S32: current measurement value at sampling instant k;

[0138] S33: current reference at sampling instant k;

[0139] S34: establishing an iterative current model;

[0140] S35: establishing a voltage vector model;

[0141] S36: Establishing a future load current model;

[0142] S37: Cost function model.

[0143] S31 load current differential model: based on sampling period T s After the load current vector dynamic vector equation is discretized, the current and voltage measurement values ​​of the current sampling period can be used to predict the load current value of the next sampling period.

[0144] The load current differential can be discretized as follows:

[0145]

[0146] Where i k is the measured current of the current sampling period.

[0147] S32 Current measurement value at sampling instant k: The current measurement value formula at sampling instant k is as follows:

[0148] i k =I meas(1) +j*I meas(2)

[0149] In the formula, I meas(1) is the transverse component of the measured current, I meas(2) To measure the longitudinal component of the current;

[0150] S33 Current reference at sampling instant k: The current reference value formula at sampling instant k is as follows:

[0151] i kref =I ref(1) +j*I ref(2)

[0152] In the formula, I ref(1) is the transverse component of the current reference value, I ref(2) is the longitudinal component of the current reference value;

[0153] S34 Iterate Current Model: Store measured current for the next iteration:

[0154] i old =i k

[0155] S35 voltage vector model: The i-th voltage vector for current prediction:

[0156] u o1 =u(i)

[0157] In the formula, the value of i ranges from 1 to 8.

[0158] S36 Future load current model: The current prediction value at time k+1 is:

[0159]

[0160] S37 Cost function model: In the MPC algorithm, for 7 different effective voltage vectors, 7 different current prediction values ​​are given based on the above formula. The corresponding voltage vector whose current prediction value is closest to the current reference value will be implemented at the next sampling moment. The basic requirement of the evaluation function is to minimize the absolute value error between the reference current and the next moment sampling current. In other words, the selected effective voltage vector will be the voltage vector that minimizes the cost function.

[0161] The cost function calculation formula is as follows:

[0162] g=|real(i kref -i k1 )|+|imag(i kref -i k1 )|

[0163] In the formula, real means taking the real part of the vector, imag means taking the imaginary part of the vector, and abs means taking the absolute value.

[0164] MPC is a model-based optimization control algorithm with inherent characteristics of adaptability and robustness. At the same time, the control objectives and constraints of the system can be included in the process of determining the system optimization performance indicators, thereby enhancing the flexibility of controller design. In the MPC controller, the load and converter models are used to predict the system behavior, so that the most appropriate output can be selected according to any set control indicators. Predictive control is a very broad concept that includes many control schemes. One of the main categories is to use predictive control to calculate the required load voltage to optimize the current behavior and use a modulator to generate the desired voltage.

[0165] Through the above method, the load current value of the next sampling period can be predicted.

[0166] Example 3

[0167] The difference between this embodiment and the above embodiment is that step S4 specifically includes the following steps:

[0168] S41: Sample the load current I generated by the voltage source inverterLOAD , the pre-selected reference current peak value I refpeak and reference frequency f ref and phase angle δ are input into the three-phase sinusoidal generator to generate the reference current I ref ;

[0169] S42:I ref and I LOAD After Clark transformation to the two-phase stationary αβ coordinate system, the reference current I ref and the measured current I meas ;

[0170] S43: The reference current I ref and the measured current I meas Bring it into the current inner loop MPC control unit and get 6 output switch states S a , S b , S c , S d , S e , S f ;

[0171] S44: According to the obtained 6 output switch states S a , S b , S c , S d , S e , S f Input to the three-phase voltage source inverter to output the most suitable voltage U abc ;

[0172] In step S43, the current inner loop MPC control process is:

[0173] S431: Obtaining the reference current value i at sampling instant k kref , which comes from the external control loop;

[0174] S432: Read the load current measurement value i at sampling instant k k ;

[0175] S433: Calculate the load back electromotive force vector e, which is used to calculate the discrete time domain model of the three-phase inverter system;

[0176] S434: Storing the measured current i for the next iteration old ;

[0177] S435: Predict the load current value i at the next sampling period, i.e., time k+1, corresponding to each different voltage vector based on the system model k1 ;

[0178] S436: Evaluate the error g between the reference current value and the predicted current value in the next sampling period based on the cost function; select the switch quantity corresponding to a vector with the smallest cost function as the input quantity of the inverter; obtain 6 output switch states S according to the optimal vector a , S b , S c , S d , S e , S f ;

[0179] S437: Outputting a voltage vector that minimizes the current error according to the six switch states as an optimal vector.

[0180] The present invention designs a novel MPC strategy. The novel MPC controller uses a discrete time domain model of a three-phase inverter system to predict the load current generated by all possible voltage vectors and selects the vector that minimizes the designed cost function as the optimal vector output.

[0181] In order to verify the effect of the present invention, simulation analysis was carried out based on the Matlab / Simulink simulation platform.

[0182] The specific parameters are as follows:

[0183] parameter Numeric Sampling period / s 0.000025 Rated frequency / Hz 50 DC voltage / V 100 Load impedance / Ω 10 Filter inductance / H 0.01 Back EMF frequency / Hz 50 Back EMF peak amplitude / V 100 Reference current peak value / A 20 Reference frequency / Hz 50

[0184] The simulation results are as follows Figure 4 , Figure 5 , Figure 6 shown. Figure 4 is the output state of the 6 switch tubes, Figure 5 This is the output load voltage A phase waveform. Figure 6 This is the waveform of the output load current phase A.

[0185] The present invention provides a model predictive control strategy based on a voltage source three-phase inverter. Based on the mathematical model of the inverter and the load, the variables can be predicted after discretization. The switch state is substituted into the designed cost function, and the optimal drive pulse can be directly output without a modulator. Compared with the traditional control strategy, this control strategy gives full play to the discrete characteristics of the inverter, has better current control dynamic performance, and is easy to implement. Thus, better dynamic and static performance is shown.

[0186] Obviously, the above embodiments are merely examples for clearly illustrating the technical solutions of the present invention, and are not intended to limit the implementation methods of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all implementation methods here. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A model predictive control strategy based on a voltage source three-phase inverter, characterized in that: The control strategy specifically includes the following steps: S1: Establish a three-phase voltage source grid-connected inverter model to obtain the voltage vector; S2: Establish a load model to determine the current vector and the back electromotive force of the load under a three-phase balanced load; S3: Build a discrete time domain model to use the current and voltage measurements of the current sampling period to predict the load current value of the next sampling period; S4: Output the most suitable voltage U abc , thus achieving predictive control.

2. The model predictive control strategy based on voltage source three-phase inverter according to claim 1, characterized in that: The step S1 comprises the following steps: S11: Establishing an inverter switch state model; S12: Establish a voltage vector model.

3. The model predictive control strategy based on voltage source three-phase inverter according to claim 2, characterized in that: The switching state of the inverter is determined by the drive signal S a , S b and S c Decide: It is expressed in vector form as follows: In the above formula, The output voltage space vector u of the inverter is: Where: u aN ,u bN and u cN is the neutral point voltage of the inverter; The relationship between vector u and S is: u=U dc WITH Where U dc is the DC voltage; Considering the driving signal S a , S b and S c A total of 8 switching states are formed, and 8 voltage vectors are obtained from them, among which the zero vector U0=U7=0, so there are 7 effective voltage vectors.

4. The model predictive control strategy based on voltage source three-phase inverter according to claim 1, characterized in that: The step S2 specifically includes the following steps: S21: Establishing a current vector model; S22: Establishing back electromotive force model; S23: Establish a load current dynamic model.

5. The model predictive control strategy based on voltage source three-phase inverter according to claim 4, characterized in that: In step S21, the current vector under three-phase balanced load is defined as: In the formula, i a 、i b 、i c They are the currents of phase a load, phase b load and phase c load respectively.

6. The model predictive control strategy based on voltage source three-phase inverter according to claim 4, characterized in that: In step S22, the back electromotive force of the load can be defined as: Where: R and L are load impedance and filter inductance; T s is the sampling time of the prediction algorithm; x old is the optimal vector at time k-1; i old is the measured current at time k-1.

7. The model predictive control strategy based on voltage source three-phase inverter according to claim 4, characterized in that: In step S23, the load current dynamics is described by the following vector equation: Where u is the inverter output voltage vector.

8. The model predictive control strategy based on voltage source three-phase inverter according to claim 1, characterized in that: The step S3 specifically comprises the following steps: S31: Establish load current differential model; S32: current measurement value at sampling instant k; S33: current reference at sampling instant k; S34: establishing an iterative current model; S35: establishing a voltage vector model; S36: Establishing a future load current model; S37: Cost function model.

9. The model predictive control strategy based on voltage source three-phase inverter according to claim 8, characterized in that: Based on the sampling period T s After the load current vector dynamic vector equation is discretized, the current and voltage measurement values ​​of the current sampling period can be used to predict the load current value of the next sampling period; The load current differential can be discretized as follows: Where i k is the measured current of the current sampling period; The formula for the current measurement value at the sampling instant k is as follows: i k =I meas(1) +j*I meas(2) In the formula, I meas(1) is the transverse component of the measured current, I meas(2) To measure the longitudinal component of the current; The current reference value formula at the sampling moment k is as follows: i kref =I ref(1) +j*I ref(2) In the formula, I ref(1) is the transverse component of the current reference value, I ref(2) is the longitudinal component of the current reference value; The iterative current model stores the measured current for the next iteration: i old =i k The voltage vector model is used for the i-th voltage vector of current prediction: in o1 =in(i) In the formula, the value of i ranges from 1 to 8; The current forecast value at time k+1 is calculated using the future load current model: In the MPC algorithm, for 7 different effective voltage vectors, 7 different current prediction values ​​are given; and the corresponding voltage vector whose current prediction value is closest to the current reference value will be implemented at the next sampling moment; the basic requirement of the evaluation function is to minimize the absolute value error between the reference current and the next moment sampling current; the selected effective voltage vector will be the voltage vector that minimizes the cost function; The cost function calculation formula is as follows: g=|real(i kref -i k1 )|+|imag(i kref -i k1 )| In the formula, real means taking the real part of the vector, imag means taking the imaginary part of the vector; abs means taking the absolute value.

10. The model predictive control strategy based on voltage source three-phase inverter according to claim 1, characterized in that: The step S4 specifically comprises the following steps: S41: Sample the load current I generated by the voltage source inverter LOAD , the pre-selected reference current peak value I refpeak and reference frequency f ref and phase angle δ are input into the three-phase sinusoidal generator to generate the reference current I ref ; S42:I ref and I LOAD After Clark transformation to the two-phase stationary αβ coordinate system, the reference current I ref and the measured current I meas ; S43: The reference current I ref and the measured current I meas Bring it into the current inner loop MPC control unit and get 6 output switch states S a , S b , S c , S d , S e , S f ; S44: According to the obtained 6 output switch states S a , S b , S c , S d , S e , S f Input to the three-phase voltage source inverter to output the most suitable voltage U abc ; In step S43, the current inner loop MPC control process is: S431: Obtaining the reference current value i at sampling instant k kref , which comes from the external control loop; S432: Read the load current measurement value i at sampling instant k k ; S433: Calculate the load back electromotive force vector e, which is used to calculate the discrete time domain model of the three-phase inverter system; S434: Storing the measured current i for the next iteration old ; S435: Predict the load current value i at the next sampling period, i.e., time k+1, corresponding to each different voltage vector based on the system model k1 ; S436: Evaluate the error g between the reference current value and the predicted current value in the next sampling period based on the cost function; select the switch quantity corresponding to a vector with the smallest cost function as the input quantity of the inverter; obtain 6 output switch states S according to the optimal vector a , S b , S c , S d , S e , S f ; S437: Outputting a voltage vector that minimizes the current error according to the six switch states as an optimal vector.

Images