A switchable predictive control method for active power filters

Abstract

The present invention discloses a switchable predictive control method for active power filters. This method is applicable to APF control strategies that are difficult to balance dynamic and steady-state performance, and facilitates switchable predictive control of APFs. The method designs a Model Predictive Control + Repetitive Control (MPC+RC) module with an additional predictive function. This module is combined with a Finite Control Set Model Predictive Control (FCS-MPC) module to achieve time-sharing output and real-time switching between two different predictive control strategies, thereby achieving the "fast response - precise control" optimization effect of the APF.

Description

Technical Field

[0001] The present invention relates to the technical field of active power filter control, and in particular to a switchable predictive control method applied to an active power filter. Background Art

[0002] With the addition of nonlinear loads, impact loads, and renewable energy power generation systems, power quality has been severely impacted. As a power quality management device, the Active Power Filter (APF) can compensate for the nonlinear currents of grid loads, thereby reducing the harmonic content of the grid current and improving the grid's current quality.

[0003] However, traditional control strategies applied to APFs are based on linear control theory, resulting in good steady-state performance but poor dynamic performance. In contrast, model predictive control strategies can improve dynamic performance, but they rely on precise parameter data of the physical system in the APF, making it difficult to accurately model the disturbance link, resulting in poor steady-state performance.

[0004] In a high-performance APF control system, both dynamic and steady-state performance must be balanced. To this end, two control modules with different control structures are constructed, each with its own specific performance optimization. Furthermore, an arbitration module is designed to achieve time-sharing output of control signals from the two control modules, providing a new approach to balancing the dynamic and steady-state performance of the APF. Summary of the Invention

[0005] In order to overcome the shortcomings of the existing technology, the present invention proposes a switchable predictive control method for active power filters. The control strategy applied to APF by this method is difficult to take into account both dynamic performance and steady-state performance. The switchable predictive control method applied to APF is designed. A repetitive control Model Predictive Control + Repetitive Control, MPC+RC module with additional prediction function is designed. On this basis, it is combined with the Finite Control Set Model Predictive Control, FCS-MPC module to realize time-sharing output and real-time switching of two different predictive control strategies, thereby achieving the optimization effect of "fast response-precise control" of APF.

[0006] A switchable predictive control method for an active power filter according to the present invention comprises the following steps:

[0007] S1. Obtain real-time detection signals of the power grid and nonlinear loads through the APF main circuit, including the load current value i la (k), ilb (k), i lc (k), AC voltage value e ab (k), e bc (k), inverter current value i ca (k), i cb (k), i cc (k), and the DC voltage value u dc (k);

[0008] S2. The detection signal is input to the APF controller, and the data is processed by the auxiliary module to generate the current compensation reference value, current compensation value and grid voltage value in the dq axis coordinate system and the α-β axis coordinate system;

[0009] S3. Input the processed signals into the MPC+RC module and the FCS-MPC module respectively;

[0010] S4. The MPC+RC module generates a first switching state based on the signal in the dq axis coordinate system by combining repetitive control with predictive control.

[0011] S5. The FCS-MPC module generates the second switching state through finite set model predictive control based on the signal in the α-β axis coordinate system;

[0012] S6. Perform a multi-dimensional hysteresis comparison on the dynamic error between the current compensation reference value and the current compensation value through the arbitration module, and output a reset signal RST and a selection signal SEL;

[0013] S7. The switching module selects the first switching state or the second switching state in a time-sharing manner according to the selection signal SEL, and outputs the selected state to the inverter of the APF main circuit to achieve optimized switching between dynamic performance and steady-state performance.

[0014] Preferably, the auxiliary module includes: a DC voltage control module, a current compensation value calculation module and a grid angle calculation module, wherein:

[0015] DC voltage control module, used to control the DC voltage value u dc (k) generating a d-axis current reference value;

[0016] A current compensation value calculation module is used to generate a current compensation reference value and a current compensation value in a dq-axis coordinate system and an α-β-axis coordinate system by transforming a stationary coordinate system and a rotating coordinate system;

[0017] The grid angle calculation module is used to calculate the grid angle according to the AC voltage value and output it to the current compensation value calculation module.

[0018] Preferably, the MPC+RC module includes: a d-axis control unit and a q-axis control unit, and the MPC+RC module has a built-in modulation algorithm, wherein:

[0019] The d-axis control unit and the q-axis control unit generate d-axis and q-axis voltage compensation output values ​​through a repeated control structure with an additional prediction function; the modulation algorithm is used to convert the voltage compensation output values ​​into a first switching state.

[0020] Preferably, the FCS-MPC module includes: an FCS prediction unit, a value function unit, and an optimization unit, wherein:

[0021] FCS prediction unit, generates current compensation prediction value in α-β axis coordinate system;

[0022] A value function unit calculates the value function result corresponding to each candidate voltage vector;

[0023] The optimization unit selects the switch state corresponding to the candidate voltage vector that minimizes the cost function as the second switch state.

[0024] The preferred arbitration module includes multiple submodules, specifically: arbitration submodule a, arbitration submodule b, arbitration submodule c, arbitration submodule d, arbitration submodule e, and arbitration submodule f. The outputs of all arbitration submodules are subjected to an OR operation logic, and the signal output value of at least one arbitration submodule being 1 is used as a trigger condition for switching the control signal.

[0025] The input of arbitration submodule a is the d-axis current compensation reference value at time kT , calculate N a The rate of change during time period T , the calculation formula is:

[0026] ;

[0027] Where N a is an integer greater than 0, set by the user; T is the control period of the arbitration module; k is an integer;

[0028] Input to the hysteresis comparator, output Fa signal, the value of the signal is 0 or 1;

[0029] in, 、 The lower and upper limits of the hysteresis comparator are set by the user.

[0030] The working principle of the hysteresis comparator is: When Fa=0; Based on the above, if Gradually increases, Fa remains at 0 until When Fa=1; if and Gradually decreases, Fa remains at 1 until When, Fa=0;

[0031] The input of arbitration submodule b is the q-axis current compensation reference value at time kT , calculate N b The rate of change during time period T , the calculation formula is:

[0032] ;

[0033] Where N b It is an integer greater than 0 and is set by the user;

[0034] Input to the hysteresis comparator, output Fb signal, the value of the signal is 0 or 1;

[0035] in, The lower and upper limits of the hysteresis comparator are set by the user.

[0036] The input of arbitration submodule c is the d-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT ,calculate ,have:

[0037] ;

[0038] Calculate N c The rate of change during time period T , the calculation formula is:

[0039] ;

[0040] Where N c It is an integer greater than 0 and is set by the user;

[0041] Input to the hysteresis comparator, output Fc signal, the value of the signal is 0 or 1;

[0042] in, The lower and upper limits of the hysteresis comparator are set by the user.

[0043] The input of arbitration submodule d is the d-axis current compensation value i at time kT cd (k), d-axis current compensation reference value at time kT ,calculate ,have:

[0044] ;

[0045] Input to the hysteresis comparator, output Fd signal, the value of the signal is 0 or 1;

[0046] in, 、 The lower and upper limits of the hysteresis comparator are set by the user.

[0047] The input of arbitration submodule e is the q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT ,calculate ,have:

[0048] ;

[0049] Input to the hysteresis comparator, output Fe signal, the value of the signal is 0 or 1;

[0050] in, 、 The lower and upper limits of the hysteresis comparator are set by the user.

[0051] The input of the arbitration submodule f is the d-axis current compensation value i at time kT cd (k), d-axis current compensation reference value at time kT , q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT ,calculate ,have:

[0052] ;

[0053] Input to the hysteresis comparator, output Ff signal, the value of the signal is 0 or 1;

[0054] in, 、 The lower and upper limits of the hysteresis comparator are set by the user.

[0055] The arbitration module also outputs a reset signal RST, which has two states: 0 and 1, which are set by the user program.

[0056] Preferably, the control period T of the MPC+RC module, arbitration module, FCS-MPC module, switching module and auxiliary module is independently set, including two working conditions:

[0057] When the control period T of the MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary module are all the same value, variables can be directly transferred between the modules;

[0058] When the control period T of the MPC+RC module, arbitration module, FCS-MPC module, switching module and auxiliary module is set to different values, a rate conversion module needs to be added during the variable transmission between the modules to exchange variables between the modules.

[0059] Preferably, the d-axis control unit and the q-axis control unit include a composite structure consisting of a prediction algorithm and a repetitive control algorithm to take into account both dynamic control performance and steady-state control performance.

[0060] Preferably, the generation of candidate voltage vectors follows a predefined switch state table, and the candidate voltage vectors participate in the calculation of the current compensation prediction value, and the cost function is based on the error between the α-β axis current compensation reference value and the prediction value.

[0061] Preferably, the selection signal SEL output by the arbitration module preferentially selects the first switching state of the MPC+RC module in a steady state, and switches to the second switching state of the FCS-MPC module when the dynamic load changes.

[0062] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0063] The present invention discloses a switchable predictive control method for an active power filter, which designs two predictive control modules that work simultaneously, including: 1. a repetitive control model predictive control + repetitive control (MPC + RC) module with an additional predictive function; and 2. a finite set model predictive control (FCS-MPC) module. The above two modules, through an arbitration module and a switching module, realize the time-sharing output control signals of the two control modules. The strategy proposed in the present invention optimizes the algorithm architecture in the "time dimension", has the characteristics of variable period, variable structure, and variable switching frequency, and enhances the flexibility of the algorithm. At the same time, the invention designs an arbitration mechanism and a switching mechanism, which, in the process of realizing real-time switching of the algorithm, avoids the drastic fluctuation of the load current caused by the switching process. The control strategy proposed in the present invention can optimize the current harmonic compensation effect of the APF in steady state, reduce the current fluctuation amplitude during the startup process of the APF, and improve the dynamic response speed of the APF under load changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 Schematic diagram of the implementation process of the control algorithm in the present invention;

[0065] Figure 2 Schematic diagram of the structure of the auxiliary module in the present invention;

[0066] Figure 3 Schematic diagram of the structure of the d-axis control unit in the present invention;

[0067] Figure 4 Schematic diagram of the structure of the q-axis control unit in the present invention;

[0068] Figure 5 Schematic diagram of the structure of the FCS-MPC module in the present invention;

[0069] Figure 6 Schematic diagram of the structure of the arbitration module in the present invention;

[0070] Figures 7 to 12 The hysteresis comparator of the present invention has output signals Fa, Fb, Fc, Fd, Fe, and Ff respectively.

[0071] In the picture:

[0072] 1-APF controller;

[0073] 1.1-MPC+RC module; 1.2-Arbitration module; 1.3-FCS-MPC module; 1.4-Switching module; 1.5-Auxiliary module;

[0074] 1.1ad axis control unit; 1.1bq axis control unit; 1.1c- modulation algorithm;

[0075] 1.2a-Arbitration submodule a; 1.2b-Arbitration submodule b; 1.2c-Arbitration submodule c; 1.2d-Arbitration submodule d; 1.2e-Arbitration submodule e; 1.2f-Arbitration submodule f;

[0076] 1.3a-FCS prediction unit; 1.3b-value function unit; 1.3c-optimization unit;

[0077] 1.5a-DC voltage control module; 1.5b-current compensation value calculation module; 1.5c-grid angle calculation module

[0078] 2-APF main circuit;

[0079] 2.1-Load current detection module; 2.2-AC voltage detection module; 2.3-Inverter current detection module; 2.4-Filter inductor; 2.5-Inverter; 2.6-DC voltage detection module;

[0080] 3-Power grid;

[0081] 4- non-linear load;

[0082] 5.1-unit d1; 5.2-unit d2; 5.3-unit d3; 5.4-unit d4; 5.5-d-axis output unit;

[0083] 6.1-unit q1; 6.2-unit q2; 6.3-unit q3; 6.4-unit q4; 6.5-q-axis output unit. DETAILED DESCRIPTION

[0084] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0085] A switchable predictive control method for an active power filter according to the present invention is specifically as follows:

[0086] The modules contained in the APF controller include: MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary module;

[0087] The auxiliary module receives the load current value i at time kT la (k), i lb (k), i lc (k), the AC voltage value e at time kT ab (k), e bc (k), the inverter current value i at time kT ca (k), i cb (k), i cc (k), DC voltage value u at time kT dc (k), after data processing, the d-axis AC voltage value e at time kT is output d (k), q-axis AC voltage value e at time kT q (k), d-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT , d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k), DC voltage value u at time kT dc (k) to MPC+RC module;

[0088] At the same time, the auxiliary module outputs the α-axis AC voltage value e at time kT α (k), β-axis AC voltage value e at time kT β (k), α-axis current compensation reference value at time kT , β-axis current compensation reference value at time kT , α-axis current compensation value i at time kT cα (k), β-axis current compensation value i at time kT cβ (k), DC voltage value u at time kT dc (k) to the FCS-MPC module, and output the d-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT , d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k) to the arbitration module; k is an integer;

[0089] The above T is the control period of the auxiliary module, which may be consistent with or inconsistent with the period of the sampling and quantization links.

[0090] The switching module receives the first switching state output by the MPC+RC module, the reset signal RST and the selection signal SEL output by the arbitration module, and the second switching state output by the FCS-MPC module, and after judgment, outputs the final switching state to the inverter.

[0091] The MPC+RC module includes a d-axis control unit, a q-axis control unit, and a modulation algorithm; wherein the d-axis control unit and the q-axis control unit respectively output the d-axis voltage compensation output value at time kT. , q-axis voltage compensation output value at time kT To the modulation algorithm; the modulation algorithm can adopt the modulation algorithm such as space vector pulse width modulation, sine pulse width modulation, etc., and is not limited to one modulation algorithm; after the modulation algorithm is processed, the first switch state is output to the switching module; see the attached Figure 1 ;

[0092] The d-axis control unit includes: unit d1, unit d2, unit d3, unit d4, and a d-axis output unit, wherein:

[0093] The input of unit d1 is the d-axis current compensation value i at time kT cd (k), d-axis AC voltage value e at time kT d (k), q-axis current compensation value i at time kT cq D-axis voltage compensation output value at time (k) and (k-1)T T is the control period of the MPC+RC module, and k is an integer. The control period of the MPC+RC module can be consistent with or inconsistent with the period of the sampling and quantization links. The output of unit d1 is the d-axis I-type current correction value at time kT. The relationship between the above variables is expressed as follows: d-axis I-type current correction formula:

[0094] ;

[0095] Where k c1 The user sets the variable in the range of 0~1, ω=2πf, f is the grid frequency, and L is the inductance of the filter inductor;

[0096] The input of unit d2 is the d-axis current compensation reference value at time kT , d-axis AC voltage value e at time kT d (k), q-axis current compensation value i at time kT cq (k), d-axis I-type current correction value at time kT , the output is p d (k), and The relationship between the above variables is expressed as follows: d-axis repetitive control structure:

[0097] ;

[0098] Where y d (k), y d (kM), y d (k+DM) are intermediate variables, and their initial values ​​are set by the user; C z , Q z Variables set by the user, all positive numbers;

[0099] In unit d2, the function of delay link Dd2 is to delay the input by MT time and then output it. The input is y d (k), the output is y d (kM); where M=1 / (fT); f is the grid frequency;

[0100] In unit d2, the function of delay link Dd1 is to output the input after delaying (MD)T time, that is, the input is y d (k), the output is y d (k+DM), the setting value of D is determined by the user and the value range is 0~M;

[0101] The input to unit d3 is , the output is h d (k); The relationship between the above variables is expressed as:

[0102] ;

[0103] Where h dp (k), z d (k), x d(k), z d (k-1), x d (k-1) are all intermediate variables, and their initial values ​​are set by the user; C r 、k p 、k i Variables set by the user, all positive numbers;

[0104] In unit d3, the function of delay link Dd3 and delay link Dd5 is to delay the input value by T time before outputting it, that is, the input value is x d (k), the output is x d (k-1); the input is z d (k), the output is z d (k-1);

[0105] In unit d3, the limiter Ld1 limits the input variable, and the relationship is:

[0106] ;

[0107] Where maxh d 、minh d The user sets the variables, which are the maximum and minimum values ​​of the output of the limiter Ld1;

[0108] The input of unit d4 is the d-axis AC voltage value e at time kT d (k), d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k), d-axis current compensation reference value at time kT , d-axis voltage compensation output value at time (k-1)T , DC voltage value u at time kT dc (k), the output is the d-axis voltage compensation prediction value u at time kT cd (k); The relationship between the above variables is expressed as:

[0109] ;

[0110] Where, is the d-axis type II current correction value at time kT, k c2 Set variables for users, ranging from 0 to 1;

[0111] The inputs of the d-axis output unit are the p output of unit d2 d (k), h output by unit d3 d (k), u output by unit d4 cd (k); The output of the d-axis output unit is the d-axis voltage compensation output value at time kT ; The relationship is:

[0112] ;

[0113] Where, is an intermediate variable, maxu cd 、minu cd k is a user-set variable, which is the maximum and minimum output value of limiter Ld2; x A variable is set for the user, ranging from 0 to 1. The value can change over time, and the change curve can be set by the user;

[0114] In the d-axis output unit, the function of the delay link Dd4 is to delay the input value by T time before outputting it, that is, the input value is When the output is ;

[0115] The q-axis control unit includes: unit q1, unit q2, unit q3, unit q4, and a q-axis output unit;

[0116] The input of unit q1 is the q-axis current compensation value i at time kT cq (k), q-axis AC voltage value e at time kT q (k), d-axis current compensation value i at time kT cd The q-axis voltage compensation output value at time (k) and (k-1)T , output the q-axis I-type current correction value at time kT The relationship between the above variables is expressed as follows: q-axis I-type current correction formula:

[0117] ;

[0118] Where k c1 Set variables for users, ranging from 0 to 1;

[0119] The input of unit q2 is the q-axis current compensation reference value at time kT , q-axis AC voltage value e at time kT q (k), d-axis current compensation value i at time kT cd (k), q-axis I-type current correction value at time kT , the output is p q (k), and The relationship between the above variables is expressed as follows: q-axis repetitive control structure:

[0120] ;

[0121] Where y q (k), y q (kM), yq (k+DM) are intermediate variables, and their initial values ​​are set by the user; C z , Q z Variables set by the user, all positive numbers;

[0122] In unit q2, the function of the delay link Dq2 is to delay the input by MT time and then output it. The input is y q (k), the output is y q (kM); where M=1 / (fT); f is the grid frequency;

[0123] In unit q2, the function of the delay link Dq1 is to delay the input by (MD)T time and then output it, that is, the input is y q (k), the output is y q (k+DM), the setting value of D is determined by the user and the value range is 0~M;

[0124] The input of unit q3 is , the output is h q (k); The relationship between the above variables is expressed as:

[0125] ;

[0126] Where h qp (k), z q (k), x q (k), z q (k-1), x q (k-1) are all intermediate variables, and their initial values ​​are set by the user; C r 、k p 、k i Set variables for users;

[0127] In unit q3, the function of delay link Dq3 and delay link Dq5 is to delay the input value by T time and then output it. That is, the input value is x q (k), the output is x q (k-1); the input is z q (k), the output is z q (k-1);

[0128] In unit q3, the limiter Lq1 limits the input variable, and the relationship is:

[0129] ;

[0130] Where maxh q 、minh q The user sets the variables, which are the maximum and minimum values ​​of the output of the limiter Lq1;

[0131] The input of unit q4 is the q-axis AC voltage value e at time kT q (k), d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT , q-axis voltage compensation output value at (k-1)T , DC voltage value u at time kT dc (k), the output is the q-axis voltage compensation prediction value u at time kT cq (k); The relationship between the above variables is expressed as:

[0132] ;

[0133] Where, is the q-axis type II current correction value at time kT, k c2 Set variables for users, ranging from 0 to 1;

[0134] The inputs of the q-axis output unit are the p output of unit q2 and q (k), h output by unit q3 q (k), u output by unit q4 cq (k); The output of the q-axis output unit is the q-axis voltage compensation output value at time kT ; The relationship is:

[0135] ;

[0136] Where, is an intermediate variable, maxu cq 、minu cq k is the user-set variable, which is the maximum and minimum output value of limiter Lq2; x Set variables for users, ranging from 0 to 1;

[0137] In the q-axis output unit, the function of the delay link Dq4 is to delay the input value by T time before outputting it, that is, the input value is When the output is .

[0138] The arbitration module includes: arbitration submodule a, arbitration submodule b, arbitration submodule c, arbitration submodule d, arbitration submodule e, arbitration submodule f, and an OR operation;

[0139] The input of arbitration submodule a is the d-axis current compensation reference value at time kT , calculate N a The rate of change during time period T for:

[0140] ;

[0141] Where N a is an integer greater than 0 and is set by the user; T is the control period of the arbitration module; k is an integer.

[0142] Input to the hysteresis comparator, output Fa signal, the value of the signal is 0 or 1;

[0143] The input of arbitration submodule b is the q-axis current compensation reference value at time kT , calculate N b The rate of change during time period T :

[0144] ;

[0145] Where N b It is an integer greater than 0 and is set by the user;

[0146] Input to the hysteresis comparator, output Fb signal, the value of the signal is 0 or 1;

[0147] The input of arbitration submodule c is the d-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT ,calculate :

[0148] ;

[0149] Calculate N c The rate of change during time period T :

[0150] ;

[0151] Where N c It is an integer greater than 0 and is set by the user;

[0152] Input to the hysteresis comparator, output Fc signal, the value of the signal is 0 or 1;

[0153] The input of arbitration submodule d is the d-axis current compensation value i at time kT cd (k), d-axis current compensation reference value at time kT ,calculate :

[0154] ;

[0155] Input to the hysteresis comparator, output Fd signal, the value of the signal is 0 or 1;

[0156] The input of arbitration submodule e is the q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT ,calculate :

[0157] ;

[0158] Input to the hysteresis comparator, output Fe signal, the value of the signal is 0 or 1;

[0159] The input of the arbitration submodule f is the d-axis current compensation value i at time kT cd (k), d-axis current compensation reference value at time kT , q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT ,calculate :

[0160] ;

[0161] Input to the hysteresis comparator, output Ff signal, the value of the signal is 0 or 1;

[0162] In addition, the arbitration module outputs a reset signal RST, which has two states: 0 and 1, which are set by the user program.

[0163] The switching module has the following inputs: the first switch state output by the MPC+RC module, the second switch state output by the FCS-MPC module, the reset signal RST, and the channel selection signal SEL;

[0164] When the channel selection signal SEL is 1, the first switch state is selected as the final switch state and output to the inverter; when the channel selection signal SEL is 0, the second switch state is selected as the final switch state and output to the inverter; the effect of the reset signal RST on the final switch state is set by the user program.

[0165] Each module in the APF controller is implemented as a Field-Programmable Gate Array (FPGA) or a multi-core microcontroller unit (MCU).

[0166] When implemented using FPGA, the MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary module are implemented by components such as configurable logic blocks in the FPGA;

[0167] When implemented using a multi-core microcontroller, the MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary module are implemented by each core in the multi-core microcontroller;

[0168] When implemented using an FPGA or multi-core microcontroller, each module runs simultaneously;

[0169] The control period T of each module, including the MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary module, can be set independently. There are two cases:

[0170] When the control periods T of the above modules are all the same value, variables can be directly transferred between the modules.

[0171] When the control periods T of the aforementioned modules are set to different values, a rate conversion module can be added during variable transmission between modules to enable variable exchange between them. Regarding setting the control periods T of each module, one approach is to use shorter control periods for the arbitration module, FCS-MPC module, switching module, and auxiliary module, while using longer control periods for the MPC+RC module. This allows the control strategy to simultaneously optimize steady-state control accuracy and dynamic response speed. Setting the control periods T of each module is user-configurable and is not limited to the above approach.

[0172] In order to more clearly illustrate the specific embodiment of the present invention, an embodiment is provided below:

[0173] The implementation process of this control algorithm is shown in the attached Figure 1 shown.

[0174] APF consists of two parts: APF controller and APF main loop.

[0175] The APF main circuit is connected to the power grid and non-linear loads.

[0176] The modules contained in the APF main circuit include: load current detection module, AC voltage detection module, inverter current detection module, filter inductor, inverter, and DC voltage detection module.

[0177] The modules contained in the APF controller include: MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary module.

[0178] The relationship between the APF controller and the APF main loop is as follows:

[0179] The APF main circuit outputs the measured values ​​of current and voltage to the APF controller, that is, the load current value i output by the load current detection module la 、i lb 、i lc ; The AC voltage value e output by the AC voltage detection module ab 、e bc ; The inverter current value i output by the inverter current detection module ca 、i cb 、i cc ; The DC voltage value u output by the DC voltage detection module dc R and L are the resistance and inductance of the filter inductor respectively; C is the capacitance value of the DC side of the inverter.

[0180] The APF controller processes the measured values ​​output by the above detection module and outputs the final switch state to the inverter in the APF main circuit. The inverter receives the final switch state issued by the APF controller, thereby realizing the inverter current value i ca 、i cb 、i cc The load current value i is finally realized la 、i lb 、i lc Real-time compensation.

[0181] The voltage and current input to the APF controller are sampled and quantized to obtain the load current value i at time kT. la (k), i lb (k), i lc (k), the AC voltage value e at time kT ab (k), e bc (k), the inverter current value i at time kT ca (k), i cb (k), i cc (k), DC voltage value u at time kT dc (k). T is the sampling and quantization period, and k is an integer. The above variables are further input into the various submodules of the APF controller.

[0182] The relationships among the MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary modules are as follows:

[0183] The auxiliary module receives the load current value i at time kT la (k), i lb (k), i lc (k), the AC voltage value e at time kT ab (k), e bc (k), the inverter current value i at time kTca (k), i cb (k), i cc (k), DC voltage value u at time kT dc (k), after data processing, the d-axis AC voltage value e at time kT is output d (k), q-axis AC voltage value e at time kT q (k), d-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT , d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k), DC voltage value u at time kT dc (k) To MPC+RC module.

[0184] At the same time, the auxiliary module outputs the α-axis AC voltage value e at time kT α (k), β-axis AC voltage value e at time kT β (k), α-axis current compensation reference value at time kT , β-axis current compensation reference value at time kT , α-axis current compensation value i at time kT cα (k), β-axis current compensation value i at time kT cβ (k), DC voltage value u at time kT dc (k) to the FCS-MPC module, and output the d-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT , d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k) To the arbitration module.

[0185] The above T is the control period of the auxiliary module, and k is an integer. The control period of the auxiliary module may be consistent with or inconsistent with the period of the sampling and quantization links.

[0186] The switching module receives the first switching state output by the MPC+RC module, the reset signal RST and the selection signal SEL output by the arbitration module, and the second switching state output by the FCS-MPC module, and after judgment, outputs the final switching state to the inverter.

[0187] The auxiliary module includes a DC voltage control module, a current compensation value calculation module, and a grid angle calculation module. The DC voltage control module inputs the d-axis current reference value at time kT to the current compensation value calculation module. The grid angle calculation module inputs the grid angle value θ at the time kT to the current compensation value calculation module. e (k).

[0188] The specific structure of the auxiliary module is shown in the attached Figure 2 shown.

[0189] The input of the DC voltage control module is the DC voltage value u at time kT dc (k), the output is the d-axis current reference value at time kT The relationship between the above variables is expressed as:

[0190] ;

[0191] Where, 、b Id (k), b Id (k-1), c Id (k), c Id (k-1) are all intermediate variables, and their initial values ​​are set by the user; 、 、 、 Set variables for the user.

[0192] In the DC voltage control module, the function of the delay link DDC1 and the delay link DDC2 is to delay the input value by T time and then output it. That is, the input value is c Id (k), the output is c Id (k-1); the input is b Id (k), the output is b Id (k-1);

[0193] In the DC voltage control module, the limiter LDC1 limits the input variable, and the relationship is:

[0194] ;

[0195] Where maxi d 、mini d The variables are set by the user, which are the maximum and minimum values ​​of the output of the limiter LDC1.

[0196] The input of the current compensation value calculation module is the load current value i at time kT la (k), i lb (k), i lc (k), the inverter current value i at time kT ca (k), i cb (k), i cc (k), d-axis current reference value at time kT , the grid angle value θ at time kT e (k). The output of the current compensation value calculation module is the d-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT , α-axis current compensation reference value at time kT , β-axis current compensation reference value at time kT , d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k), α-axis current compensation value i at time kT cα (k), β-axis current compensation value i at time kT cβ (k).

[0197] The inverter current value i at time kT ca (k), i cb (k), i cc (k) is input into the current static coordinate system transformation b1, and the α-axis current compensation value i at time kT is output. cα (k), β-axis current compensation value i at time kT cβ (k). The relationship is:

[0198] ;

[0199] The α-axis current compensation value i at time kT cα (k), β-axis current compensation value i at time kT cβ (k), grid angle value θ at time kT e (k) is input to the current rotation coordinate system transformation b2, and the d-axis current compensation value i at time kT is output. cd (k), q-axis current compensation value i at time kT cq (k). The relationship is:

[0200] ;

[0201] The above calculation formula obtains the α-axis current compensation value i at time kT cα (k), β-axis current compensation value i at time kT cβ (k), d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k), these variables are output from the current compensation value calculation module.

[0202] The load current value i at time kT la (k), i lb (k), i lc(k) is input into the current stationary coordinate system transformation b3, and the α-axis load current value i at time kT is output. lα (k), β-axis load current value i at time kT lβ (k). The relationship is:

[0203] ;

[0204] The α-axis load current value i at time kT lα (k), β-axis load current value i at time kT lβ (k), grid angle value θ at time kT e (k) is input to the current rotation coordinate system transformation b4, and the d-axis load current value i at time kT is output. ld (k), q-axis load current value i at time kT lq (k). The relationship is:

[0205] ;

[0206] D-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT The calculation formula is as follows:

[0207] ;

[0208] Where i ldf (k), i ldf (k-1), i lqf (k), i lqf (k-1), 、 、 are all intermediate variables, and their initial values ​​are set by the user. In addition, Always 0; 、 Set the variable for the user, ranging from 0 to 1.

[0209] In the current compensation value calculation module, the function of delay link DC2 and delay link DC1 is to delay the input value by T time and then output it. That is, the input value is i ldf (k), the output is i ldf (k-1); the input is i lqf (k), the output is i lqf (k-1); that is, the input is When the output is ;

[0210] D-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT , the grid angle value θ at time kT e (k) is input into the inverse transformation b5 of the current rotating coordinate system, and the α-axis current compensation reference value at time kT is output. , β-axis current compensation reference value at time kT The relationship is:

[0211] ;

[0212] The above calculation formula obtains the d-axis current compensation reference value at time kT: , q-axis current compensation reference value at time kT , α-axis current compensation reference value at time kT , β-axis current compensation reference value at time kT , these variables are output from the current compensation value calculation module.

[0213] The input of the grid angle calculation module is the AC voltage value e at time kT ab (k), e bc (k), after voltage stationary coordinate system transformation c1, output the α-axis AC voltage value e at time kT α (k), β-axis AC voltage value e at time kT β (k). The relationship is:

[0214] ;

[0215] The α-axis AC voltage value e at time kT α (k), β-axis AC voltage value e at time kT β (k), θ frac (k-1), after the voltage rotation coordinate system transformation c2, the d-axis AC voltage value e at time kT is output d (k), q-axis AC voltage value e at time kT q (k). The relationship is:

[0216] ;

[0217] Where θ frac (k-1) is an intermediate variable, and its initial value is set by the user.

[0218] The q-axis AC voltage value e at time kT q (k) and the intermediate variable θ o The relationship between (k) is:

[0219] ;

[0220] Where θ q (k),θ q(k-1) is an intermediate variable, whose initial value is set by the user; 、 Set variables for the user.

[0221] In the grid angle calculation module, the function of the delay links DE1, DE2, and DE3 is to delay the input value by T time before outputting it, that is, the input value is θ q (k), the output is θ q (k-1); the input is θ FL1 (k), the output is θ FL1 (k-1); the input is θ frac (k), the output is θ frac (k-1);

[0222] In the grid angle calculation module, the limiter LE1 is used to calculate the intermediate variable θ o (k) is restricted, the relationship is:

[0223] ;

[0224] Where maxθ o 、minθ o The variables are set by the user, which are the maximum and minimum values ​​of the output of the limiter LE1.

[0225] Intermediate variable θ FL1 (k) and the intermediate variable θ LE1 The relationship between (k) is:

[0226] ;

[0227] Where, Set the variable for the user, ranging from 0 to 1.

[0228] In the grid angle calculation module, the intermediate variable θ FL1 (k) Take the decimal part and get θ frac (k), and find the absolute value of its decimal part to obtain the grid angle value θ at time kT e (k) and output it to the grid angle calculation module.

[0229] The MPC+RC module includes a d-axis control unit, a q-axis control unit, and a modulation algorithm. The d-axis control unit and the q-axis control unit each output the d-axis voltage compensation output value at time kT. , q-axis voltage compensation output value at time kT To the modulation algorithm. The modulation algorithm can adopt the modulation algorithm such as space vector pulse width modulation, sine pulse width modulation, etc., and is not limited to one modulation algorithm. After the modulation algorithm is processed, the switch state 1 is output to the switching module. See the attached Figure 1 .

[0230] The specific structure of the d-axis control unit is shown in the attached Figure 3 shown.

[0231] The d-axis control unit consists of unit d1, unit d2, unit d3, unit d4, and d-axis output unit. The input of unit d1 is the d-axis current compensation value i at time kT. cd (k), d-axis AC voltage value e at time kT d (k), q-axis current compensation value i at time kT cq D-axis voltage compensation output value at time (k) and (k-1)T T is the control period of the MPC+RC module, and k is an integer. The output of unit d1 is the d-axis I-type current correction value at time kT. The relationship between the above variables is expressed as: d-axis I-type current correction form, specifically:

[0232] ;

[0233] Where k c1 It is a user-set variable ranging from 0 to 1, ω=2πf, and f is the grid frequency.

[0234] The input of unit d2 is the d-axis current compensation reference value at time kT , d-axis AC voltage value e at time kT d (k), q-axis current compensation value i at time kT cq (k), d-axis I-type current correction value at time kT , the output is p d (k), and The relationship between the above variables is expressed as: d-axis repetitive control structure, specifically:

[0235] ;

[0236] Where y d (k), y d (kM), y d (k+DM) are intermediate variables, and their initial values ​​are set by the user; C z , Q z Variables set by the user, all are positive numbers.

[0237] In unit d2, the function of delay link Dd2 is to delay the input by MT time and then output it. The input is y d (k), the output is y d (kM). Where M=1 / (f T); f is the grid frequency.

[0238] In unit d2, the function of delay link Dd1 is to output the input after delaying (MD)T time, that is, the input is y d (k), the output is y d (k+DM), the setting value of D is determined by the user and the value range is 0~M;

[0239] The input to unit d3 is , the output is h d (k); The relationship between the above variables is expressed as:

[0240] ;

[0241] Where h dp (k), z d (k), x d (k), z d (k-1), x d (k-1) are all intermediate variables, and their initial values ​​are set by the user; C r 、k p 、k i Variables set by the user, all are positive numbers.

[0242] In unit d3, the function of delay link Dd3 and delay link Dd5 is to delay the input value by T time before outputting it, that is, the input value is x d (k), the output is x d (k-1); the input is z d (k), the output is z d (k-1);

[0243] In unit d3, the limiter Ld1 limits the input variable, and the relationship is:

[0244] ;

[0245] Where maxh d 、minh d The variables are set by the user, which are the maximum and minimum values ​​of the output of the limiter Ld1.

[0246] The input of unit d4 is the d-axis AC voltage value e at time kT d (k), d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k), d-axis current compensation reference value at time kT , d-axis voltage compensation output value at time (k-1)T , DC voltage value u at time kT dc (k), the output is the d-axis voltage compensation prediction value u at time kTcd (k); The relationship between the above variables is expressed as:

[0247] ;

[0248] Where, is the d-axis type II current correction value at time kT, k c2 Set the variable for the user, ranging from 0 to 1.

[0249] The inputs of the d-axis output unit are the p output of unit d2 d (k), h output by unit d3 d (k), u output by unit d4 cd (k). The output of the d-axis output unit is the d-axis voltage compensation output value at time kT. The relationship is:

[0250] ;

[0251] ;

[0252] Where, is an intermediate variable, maxu cd 、minu cd k is a user-set variable, which is the maximum and minimum output value of limiter Ld2; x The variable is set by the user, ranging from 0 to 1. The value can change over time, and the change curve can be set by the user.

[0253] In the d-axis output unit, the function of the delay link Dd4 is to delay the input value by T time before outputting it, that is, the input value is When the output is .

[0254] The specific structure of the q-axis control unit is shown in the attached Figure 4 shown.

[0255] The q-axis control unit consists of unit q1, unit q2, unit q3, unit q4, and a q-axis output unit.

[0256] The input of unit q1 is the q-axis current compensation value i at time kT cq (k), q-axis AC voltage value e at time kT q (k), d-axis current compensation value i at time kT cd The q-axis voltage compensation output value at time (k) and (k-1)T , output the q-axis I-type current correction value at time kT The relationship between the above variables is expressed as: q-axis I-type current correction form, specifically:

[0257] ;

[0258] Where k c1 Set the variable for the user, ranging from 0 to 1.

[0259] The input of unit q2 is the q-axis current compensation reference value at time kT , q-axis AC voltage value e at time kT q (k), d-axis current compensation value i at time kT cd (k), q-axis I-type current correction value at time kT , the output is p q (k), and The relationship between the above variables is expressed as follows: q-axis repetitive control structure:

[0260] ;

[0261] Where y q (k), y q (kM), y q (k+DM) are intermediate variables, and their initial values ​​are set by the user; C z , Q z Variables set by the user, all are positive numbers.

[0262] In unit q2, the function of the delay link Dq2 is to delay the input by MT time and then output it. The input is y q (k), the output is y q (kM). Where M=1 / (f T); f is the grid frequency.

[0263] In unit q2, the function of the delay link Dq1 is to delay the input by (MD)T time and then output it, that is, the input is y q (k), the output is y q (k+DM), the setting value of D is determined by the user and the value range is 0~M;

[0264] The input of unit q3 is , the output is h q (k); The relationship between the above variables is expressed as:

[0265] ;

[0266] ;

[0267] Where h qp (k), z q (k), x q (k), z q (k-1), xq (k-1) are all intermediate variables, and their initial values ​​are set by the user; C r 、k p 、k i Set variables for the user.

[0268] In unit q3, the function of delay link Dq3 and delay link Dq5 is to delay the input value by T time and then output it. That is, the input value is x q (k), the output is x q (k-1); the input is z q (k), the output is z q (k-1);

[0269] In unit q3, the limiter Lq1 limits the input variable, and the relationship is:

[0270] ;

[0271] Where maxh q 、minh q The variables are set by the user, which are the maximum and minimum values ​​of the output of the limiter Lq1.

[0272] The input of unit q4 is the q-axis AC voltage value e at time kT q (k), d-axis current compensation value i at time kT cd (k), q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT , q-axis voltage compensation output value at (k-1)T , DC voltage value u at time kT dc (k), the output is the q-axis voltage compensation prediction value u at time kT cq (k); The relationship between the above variables is expressed as:

[0273] ;

[0274] Where, is the q-axis type II current correction value at time kT, k c2 Set the variable for the user, ranging from 0 to 1.

[0275] The inputs of the q-axis output unit are the p output of unit q2 and q (k), h output by unit q3 q (k), u output by unit q4 cq (k). The output of the q-axis output unit is the q-axis voltage compensation output value at time kT. The relationship is:

[0276] ;

[0277] Where, is an intermediate variable, maxu cq 、minu cq k is the user-set variable, which is the maximum and minimum output value of limiter Lq2; x Set the variable for the user, ranging from 0 to 1.

[0278] In the q-axis output unit, the function of the delay link Dq4 is to delay the input value by T time before outputting it, that is, the input value is When the output is .

[0279] The FCS-MPC module includes FCS prediction unit, value function unit, and optimization unit. The specific structure is shown in the attached Figure 5 shown.

[0280] The inputs of the FCS prediction unit are the α-axis AC voltage value e at time kT and α (k), β-axis AC voltage value e at time kT β (k), α-axis current compensation value i at time kT cα (k), β-axis current compensation value i at time kT cβ (k), DC voltage value u at time kT dc (k). T is the control period of the FCS-MPC module, and k is an integer.

[0281] Among them, the DC voltage value u at time kT is dc (k) After the voltage vector generation module, the α-axis voltage candidate value is output , β-axis voltage candidate value .

[0282] Furthermore, the α-axis current compensation prediction value at time k+1 is obtained by the following calculation formula: , β-axis current compensation prediction value at time k+1 ,have:

[0283] ;

[0284] The above formula will be calculated 8 times. In each calculation, the value of n increases gradually from 0 to 7.

[0285] The voltage vector generation follows the table below

[0286]

[0287] The input of the value function unit is the α-axis current compensation prediction value at time k+1 , β-axis current compensation prediction value at time k+1 , α-axis current compensation reference value at time kT , β-axis current compensation reference value at time kT Among them, there are , , the output is the value function calculation result , the calculation formula is:

[0288] ;

[0289] The above formula will be calculated 8 times, and in each calculation, the value of n increases gradually from 0 to 7.

[0290] The input of the optimization unit is , through Compare and take the n value corresponding to the minimum value and assign it to n o , and n o The switching state corresponding to the value is output to the inverter. The corresponding switching state is named switching state 2. The calculation formula is:

[0291] ;

[0292] n o The relationship between the value and switch state 2 is shown in the following table

[0293]

[0294] In the above table, sw a , sw b , sw c These represent the switching states of the upper-arm power devices in the inverter's phases a, b, and c, respectively. 1 represents on, and 0 represents off. For any arm, the switching state of its lower arm is the opposite of that of its upper arm.

[0295] Arbitration module as attached Figure 6 As shown, it includes arbitration submodule a, arbitration submodule b, arbitration submodule c, arbitration submodule d, arbitration submodule e, arbitration submodule f, and an OR operation ( Figure 6 in OR).

[0296] The input of arbitration submodule a is the d-axis current compensation reference value at time kT , calculate N a The rate of change during time period T , the calculation formula is:

[0297] ;

[0298] Where N ais an integer greater than 0, set by the user; T is the control period of the arbitration module; k is an integer;

[0299] Input to the attached Figure 7 The hysteresis comparator outputs the Fa signal, which has a value of 0 or 1.

[0300] In the figure, 、 The lower and upper limits of the hysteresis comparator are set by the user.

[0301] The working principle of the hysteresis comparator is: When Fa=0; Based on the above, if Gradually increases, Fa remains at 0 until When Fa=1; if and Gradually decreases, Fa remains at 1 until When, Fa=0.

[0302] The principle of the hysteresis comparator described below is the same as that described above.

[0303] The input of arbitration submodule b is the q-axis current compensation reference value at time kT , calculate N b The rate of change during time period T , the calculation formula is:

[0304] ;

[0305] Where N b It is an integer greater than 0 and is set by the user;

[0306] Input to the attached Figure 8 The hysteresis comparator outputs the Fb signal, which has a value of 0 or 1.

[0307] In the figure, 、 The lower and upper limits of the hysteresis comparator are set by the user.

[0308] The input of arbitration submodule c is the d-axis current compensation reference value at time kT , q-axis current compensation reference value at time kT ,calculate ,have:

[0309] ;

[0310] Calculate N c The rate of change during time period T , the calculation formula is:

[0311] ;

[0312] Where N c It is an integer greater than 0 and is set by the user;

[0313] Input to the attached Figure 9 The hysteresis comparator outputs the Fc signal, which has a value of 0 or 1.

[0314] In the figure, The lower and upper limits of the hysteresis comparator are set by the user.

[0315] The input of arbitration submodule d is the d-axis current compensation value i at time kT cd (k), d-axis current compensation reference value at time kT ,calculate ,have:

[0316] ;

[0317] Input to the attached Figure 10 The hysteresis comparator outputs the Fd signal, which has a value of 0 or 1.

[0318] In the figure, The lower and upper limits of the hysteresis comparator are set by the user.

[0319] The input of arbitration submodule e is the q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT ,calculate ,have:

[0320] ;

[0321] Input to the attached Figure 11 The hysteresis comparator outputs the Fe signal, which has a value of 0 or 1.

[0322] In the figure, The lower and upper limits of the hysteresis comparator are set by the user.

[0323] The input of the arbitration submodule f is the d-axis current compensation value i at time kT cd (k), d-axis current compensation reference value at time kT , q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT ,calculate ,have:

[0324] ;

[0325] Input to the attached Figure 12 The hysteresis comparator outputs the Ff signal, which has a value of 0 or 1.

[0326] In the figure, The lower and upper limits of the hysteresis comparator are set by the user.

[0327] In addition, the arbitration module outputs a reset signal RST, which has two states: 0 and 1, which are set by the user program.

[0328] The inputs of the switching module are the switch state 1 output by the MPC+RC module, the switch state 2 output by the FCS-MPC module, the reset signal RST, and the channel selection signal SEL.

[0329] When the channel select signal SEL is 1, switch state 1 is selected as the final switch state and output to the inverter. When the channel select signal SEL is 0, switch state 2 is selected as the final switch state and output to the inverter. The effect of the reset signal RST on the final switch state is set by the user program.

[0330] The various modules in the APF controller can be implemented using a field-programmable gate array (FPGA) or a multi-core microcontroller unit (MCU). When implemented using an FPGA, the MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary modules can be implemented using components such as configurable logic blocks within the FPGA. When implemented using a multi-core microcontroller, the MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary modules can be implemented by individual cores within the multi-core microcontroller. When implemented using an FPGA or multi-core microcontroller, each module can operate simultaneously.

[0331] The control period T of each module, including the MPC+RC module, arbitration module, FCS-MPC module, switching module, and auxiliary module, can be set independently. There are two cases:

[0332] When the control periods T of the above modules are all the same value, variables can be directly transferred between the modules.

[0333] When the control periods T of the aforementioned modules are set to different values, a rate conversion module can be added during variable transmission between modules to enable variable exchange between them. Regarding setting the control periods T of each module, one approach is to use shorter control periods for the arbitration module, FCS-MPC module, switching module, and auxiliary module, while using longer control periods for the MPC+RC module. This allows the control strategy to simultaneously optimize steady-state control accuracy and dynamic response speed. Setting the control periods T of each module is user-configurable and is not limited to the above approach.

[0334] The present invention and its embodiments are described above. This description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs structures and embodiments similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. A switchable predictive control method for an active power filter, characterized by: The following steps are involved: S1. Obtain real-time detection signals of the power grid (3) and nonlinear load (4) through the APF main circuit (2), including the load current value i la (k), i lb (k), i lc (k), AC voltage value e ab (k), e bc (k), inverter current value i ca (k), i cb (k), i cc (k), and the DC voltage value u dc (k); S2. The detection signal is input to the APF controller (1), and the data is processed by the auxiliary module (1.5) to generate the current compensation reference value, the current compensation value and the grid voltage value in the dq axis coordinate system and the α-β axis coordinate system; S3. Input the processed signals into the MPC+RC module (1.1) and the FCS-MPC module (1.3) respectively; S4. The MPC+RC module (1.1) generates a first switching state based on the signal in the dq axis coordinate system by combining repetitive control with predictive control; S5. The FCS-MPC module (1.3) generates a second switching state through finite set model predictive control based on the signal in the α-β axis coordinate system; S6. Performing a multi-dimensional hysteresis comparison on the dynamic error between the current compensation reference value and the current compensation value through an arbitration module (1.2), and outputting a reset signal RST and a selection signal SEL; wherein the arbitration module (1.2) includes a plurality of submodules, specifically: an arbitration submodule a (1.2a), an arbitration submodule b (1.2b), an arbitration submodule c (1.2c), an arbitration submodule d (1.2d), an arbitration submodule e (1.2e), and an arbitration submodule f (1.2f); an OR operation logic is applied to the outputs of all arbitration submodules, and a signal output value of 1 of at least one arbitration submodule is used as a trigger condition for switching the control signal; S7. The switching module (1.4) selects the first switching state or the second switching state in a time-sharing manner according to the selection signal SEL, and outputs the selected state to the inverter (2.5) of the APF main circuit (2), thereby achieving optimized switching between dynamic performance and steady-state performance.

2. The switchable predictive control method for an active power filter according to claim 1, wherein: The auxiliary module (1.5) comprises: a DC voltage control module (1.5a), a current compensation value calculation module (1.5b) and a grid angle calculation module (1.5c), wherein: The DC voltage control module (1.5a) is used to control the DC voltage value u dc (k) generating a d-axis current reference value; The current compensation value calculation module (1.5b) is used to generate a current compensation reference value and a current compensation value in a dq-axis coordinate system and an α-β-axis coordinate system by transforming a stationary coordinate system and a rotating coordinate system; The grid angle calculation module (1.5c) is used to calculate the grid angle according to the AC voltage value and output it to the current compensation value calculation module.

3. The switchable predictive control method for an active power filter according to claim 1, wherein: The MPC+RC module (1.1) includes: a d-axis control unit (1.1a) and a q-axis control unit (1.1b), and the MPC+RC module (1.1) has a built-in modulation algorithm (1.1c), wherein: The d-axis control unit (1.1a) and the q-axis control unit (1.1b) generate d-axis and q-axis voltage compensation output values ​​through a repeated control structure with an additional prediction function; the modulation algorithm (1.1c) is used to convert the voltage compensation output value into a first switching state.

4. The switchable predictive control method for an active power filter according to claim 1, wherein: The FCS-MPC module (1.3) includes: an FCS prediction unit (1.3a), a cost function unit (1.3b) and an optimization unit (1.3c), wherein: The FCS prediction unit (1.3a) generates a current compensation prediction value in the α-β axis coordinate system; The value function unit (1.3b) calculates the value function result corresponding to each candidate voltage vector; The optimization unit (1.3c) selects the switch state corresponding to the candidate voltage vector that minimizes the cost function as the second switch state.

5. The switchable predictive control method for an active power filter according to claim 1, wherein: In the arbitration module (1.2): The input of the arbitration submodule a (1.2a) is the d-axis current compensation reference value at time kT Calculate N a The rate of change during time period T The calculation formula is: Where N a is an integer greater than 0 and is set by the user; T is the control period of the arbitration module (1.2); k is an integer; Input to the hysteresis comparator, output Fa signal, the value of the signal is 0 or 1; in, The lower and upper limits of the hysteresis comparator are set by the user. The working principle of the hysteresis comparator is: When Fa=0; Based on the above, if Gradually increases, Fa remains at 0 until When Fa=1; if and Gradually decreases, Fa remains at 1 until When, Fa=0; The input of the arbitration submodule b (1.2b) is the q-axis current compensation reference value at time kT Calculate N b The rate of change during time period T The calculation formula is: Where N b It is an integer greater than 0 and is set by the user; Input to the hysteresis comparator, output Fb signal, the value of the signal is 0 or 1; in, The lower and upper limits of the hysteresis comparator are set by the user. The input of the arbitration submodule c (1.2c) is the d-axis current compensation reference value at time kT The q-axis current compensation reference value at time kT calculate have: Calculate N c The rate of change during time period T The calculation formula is: Where N c It is an integer greater than 0 and is set by the user; Input to the hysteresis comparator, output Fc signal, the value of the signal is 0 or 1; in, The lower and upper limits of the hysteresis comparator are set by the user. The input of the arbitration submodule d(1.2d) is the d-axis current compensation value i at time kT cd (k), d-axis current compensation reference value at time kT Calculate Δi cd (k) has: Δi cd (k) Input to the hysteresis comparator, output Fd signal, the value of which is 0 or 1; Among them, Δi cdmin , Δi cdmax The lower and upper limits of the hysteresis comparator are set by the user. The input of the arbitration submodule e (1.2e) is the q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT Calculate Δi cq (k) has: Δi cq (k) Input to the hysteresis comparator, output Fe signal, the value of which is 0 or 1; Among them, Δi cqmin , Δi cqmax The lower and upper limits of the hysteresis comparator are set by the user. The input of the arbitration submodule f(1.2f) is the d-axis current compensation value i at time kT cd (k), d-axis current compensation reference value at time kT The q-axis current compensation value i at time kT cq (k), q-axis current compensation reference value at time kT Calculate Δi cdq (k) has: Δi cdq (k) Input to the hysteresis comparator, output Ff signal, the value of which is 0 or 1; Among them, Δi cdqmin , Δi cdqmax The lower and upper limits of the hysteresis comparator are set by the user. The arbitration module (1.2) also outputs a reset signal RST, which has two states, 0 and 1, and is set by a user program.

6. The switchable predictive control method for an active power filter according to claim 1, wherein: The control period T of the MPC+RC module (1.1), arbitration module (1.2), FCS-MPC module (1.3), switching module (1.4) and auxiliary module (1.5) is independently set, including two working conditions: When the control periods T of the MPC+RC module (1.1), arbitration module (1.2), FCS-MPC module (1.3), switching module (1.4) and auxiliary modules are all the same value, variables can be directly transferred between the modules; When the control periods T of the MPC+RC module (1.1), arbitration module (1.2), FCS-MPC module (1.3), switching module (1.4) and auxiliary module are set to different values, a rate conversion module needs to be added during the variable transmission process between the modules to exchange variables between the modules.

7. The switchable predictive control method for an active power filter according to claim 3, wherein: The d-axis control unit (1.1a) and the q-axis control unit (1.1b) include a composite structure consisting of a prediction algorithm and a repetitive control algorithm to take into account both dynamic control performance and steady-state control performance.

8. The switchable predictive control method for an active power filter according to claim 4, wherein: The candidate voltage vectors are generated according to a predefined switch state table, and the candidate voltage vectors participate in the calculation of the current compensation prediction value, and the cost function is based on the error between the α-β axis current compensation reference value and the prediction value.

9. A switchable predictive control method for an active power filter according to any one of claims 1 to 8, characterized in that: The selection signal SEL output by the arbitration module (1.2) preferentially selects the first switching state of the MPC+RC module (1.1) in a steady state, and switches to the second switching state of the FCS-MPC module (1.3) when the dynamic load changes.

Images