Three-phase SVPWM (Space Vector Pulse Width Modulation) grid-connected inverter harmonic control method based on improved particle swarm optimization

Abstract

The invention provides a three-phase SVPWM grid-connected inverter harmonic control method based on an improved particle swarm algorithm, and relates to the field of grid-connected control. On the basis of double closed-loop control of a three-phase SVPWM grid-connected inverter, a main circuit topological structure of the grid-connected inverter is firstly provided, and a mathematical model of the grid-connected inverter is determined; then determining a double-closed-loop control strategy based on space vector modulation according to the mathematical model; and aiming at a voltage outer loop and a current inner loop of the double-closed-loop control strategy, an improved particle swarm optimization algorithm is adopted to optimize PI parameters. According to the improved inertia weight, the minimum value, the maximum value and the fitness function of the inertia weight are comprehensively considered, a harmonic suppression performance index is added to the fitness function, and the improved particle swarm optimization algorithm enhances the global search capability and the accurate search capability, so that the probability of falling into a locally optimal solution is reduced. According to the control method, the grid-connected process of the grid-connected inverter can be more stable, the anti-interference capability is enhanced, the static error is reduced, and the electric energy quality is improved.

Description

Technical Field

[0001] The present invention relates to the field of grid-connected control, and in particular to a harmonic control method of a three-phase SVPWM grid-connected inverter based on an improved particle swarm algorithm. Background Art

[0002] At present, energy is accelerating its transformation and advancement towards green, clean and efficient directions. The grid connection of multiple energy sources faces problems such as stability, efficiency and parameter matching. The grid connection of DC power supply is inseparable from the grid-connected inverter. The grid-connected inverter is a device that converts DC power into AC power and transmits electrical energy to the public power grid. Reasonable inverter topology selection and appropriate control strategy selection directly affect the success rate and efficiency of grid connection.

[0003] According to the number of levels of the output voltage of the three-phase grid-connected inverter, the most commonly used are two-level and three-level inverters. The control strategy for these grid-connected inverters generally adopts proportional integral (PI) control. On this basis, various other control strategies are derived:

[0004] The prior art utilizes an improved feedforward compensation strategy, that is, adding a compensation unit into a feedforward loop, to solve the problem that the dynamic performance of PI control is not perfect.

[0005] The prior art also has the problem of adding fuzzy control to the PI control link of the intelligent photovoltaic grid-connected inverter to improve the problem of poor dynamic response of repeated control during the control process.

[0006] The above control method is relatively complex, and there are interferences when it is actually applied to the grid-connected inverter, and the grid-connected process is unstable. The traditional particle swarm optimization (PSO) algorithm is prone to slow convergence speed and high probability of falling into the local optimal solution when searching for the global optimal solution. Summary of the invention

[0007] Purpose of the invention: To propose a harmonic control method for a three-phase SVPWM grid-connected inverter based on an improved particle swarm algorithm to solve the above-mentioned problems existing in the prior art.

[0008] The first aspect of the present invention provides a harmonic control method for a three-phase SVPWM grid-connected inverter based on an improved particle swarm algorithm. The method comprises the following steps:

[0009] S1. Construct a main circuit model of a three-phase SVPWM grid-connected inverter, convert direct current into alternating current with the same frequency and phase as the grid voltage through the three-phase SVPWM grid-connected inverter and send it into the grid; the control loop of the three-phase SVPWM grid-connected inverter adopts double closed-loop control, including a voltage outer loop and a current inner loop;

[0010] S2. Determine the range of voltage outer loop parameters and current inner loop parameters through engineering calculations;

[0011] S3, using the improved particle swarm algorithm to find the optimal parameters within the range of the voltage outer loop parameters and the current inner loop parameters, and outputting the improved particle swarm optimization PI parameters;

[0012] S4. Use the improved particle swarm optimization PI parameters to complete the harmonic control of the three-phase SVPWM grid-connected inverter.

[0013] In a further embodiment of the first aspect, the main circuit model of the three-phase SVPWM grid-connected inverter includes a three-phase voltage-type full-bridge inverter and six power electronic switches; the power electronic switches are composed of insulated gate bipolar transistors; each power electronic switch has two states of on and off; the switch function g is defined x , x=a,b,c; when g x =1, the switches S1, S3, and S5 are turned on; when g x =0, the switches S2, S4, and S6 are turned on;

[0014] The equation in the three-phase stationary coordinate system is:

[0015]

[0016] Where U dc The DC voltage is input to the DC side; i x is the AC current at the output side of the grid-connected inverter; e x is the phase voltage of the power grid; R is the filter resistance value; L is the inductance value.

[0017] In a further embodiment of the first aspect, the process of determining the range of the voltage outer loop parameter and the current inner loop parameter by engineering calculation specifically includes:

[0018] The three-phase stationary coordinate system is converted into a vertical α-β coordinate system, and then a dq transformation is performed. The mathematical model under the dq coordinate is:

[0019]

[0020] Where, -ωLi q +ωLi q is the coupling term; U dc is the DC voltage value; i d and i q are the AC current values ​​of the power grid in dq coordinates respectively; e d and e q are the AC voltage values ​​of the power grid in dq coordinates respectively;

[0021] The current inner loop is composed of PI controller link, system delay link, inverter link, and filter link;

[0022] The voltage outer loop is referenced by the d-axis voltage u dgiven , q-axis voltage reference u qgiven The actual voltage u d 、u q Compared with PI control, the closed-loop transfer function expression is obtained:

[0023]

[0024] Where G1(s) is the transfer function of the voltage outer loop PI controller; G2(s) is the transfer function of the current inner loop; G4(s) is the feedback sampling delay;

[0025] in:

[0026]

[0027]

[0028] Where, T s is the sampling period; K Pd-in , K Id-in , K Pq-in , K Iq-i are the proportional and integral parameters of the PI controller of the d-axis and q-axis of the inner current loop respectively; K Pd-out , K Id-out , K Pq-out , K Iq-ou are the proportional and integral parameters of the PI controller of the d-axis and q-axis of the voltage outer loop respectively;

[0029] Above K Pd-in , K Id- , K Pq-i , K Iq-in , K Pd-out , K Id-out , K Pq- , K Iq-out Together they constitute the PI parameters, and the range of the PI parameters is preliminarily screened out.

[0030] In a further embodiment of the first aspect, the process of finding the optimal parameters using the improved particle swarm algorithm and outputting the improved particle swarm optimization PI parameters specifically includes:

[0031] Set the particle position, particle flight speed, number of iterations, and inertia weight, and find the PI parameters that meet the predetermined conditions through iterative update optimization; the PI parameters are replaced by the position of the particle group;

[0032] During the entire optimization process, the fitness function and the improved inertia weight affect the final result of the optimization.

[0033] In a further embodiment of the first aspect, an N-dimensional search space R N In the example, the population consists of m particles, and the particle positions are expressed as:

[0034] X i ={X i1 ,X i2 ,…X ik ,…X iN}

[0035] Where, X iN Represents the position of the i-th particle in the N-dimensional search space;

[0036] The particle flight speed is expressed as:

[0037] V i = {V i1 ,V i2 ,…V ik ,…V iN}

[0038] Initialize a group of random particles in the N-dimensional search space, and then find the optimal solution through iterative updates. In each iterative update, search for its own individual extreme value P ifit and the global extremum G ifit To update its position and flight speed;

[0039] Where P ifit Refers to the current optimal solution found by the particle itself, G ifit Refers to the optimal solution that the entire population can find; P ifit and G ifit Respectively expressed as:

[0040] P ifit = {P ifit1 ,P ifit2 ,…P ifitk ,…P ifit2}

[0041] G ifit = {G ifit1 ,G ifit2 ,…G ifitk ,…G ifitN}

[0042] According to the requirements of iterative update, the formula for particle update speed and position is:

[0043]

[0044] In the formula, in is a particle in the N-dimensional search space; k is the current iteration number; k+1 is the update iteration number; ω is the inertia weight coefficient; c1 and c2 are acceleration coefficients, representing the particle's movement toward its own individual extreme value P ifit and the global extremum G ifit The acceleration weight of the propulsion; r1 and r2 are random numbers selected between 0 and 1.

[0045] The expression of inertia weight is as follows:

[0046]

[0047] In the formula, ω max and ω min To improve the minimum and maximum values ​​of the inertia weight ω, it takes values ​​between 0 and 1; J k is the current fitness function; J min and J pj is the minimum and average value of the fitness function; ω (k) As the number of iterations and the fitness function change, P ifit >G ifit , give greater inertia weight.

[0048] Fitness function J k The expression is as follows:

[0049] J k =h1∫0 M t(|e ud (t)|+|e uq (t)|)dt+h2∫0 M t(|e id (t)|+|e iq (t)|)dt+h3∫0 M THD u dt+h4∫0 M THD i dt

[0050] In the formula, the first and second integral terms are to ensure the response speed and stability of the system; M is the data length; e ud (t), e uq (t), e id (t), e iq (t) is the difference between the given voltage and current of the d-axis and q-axis and the actual voltage and current of the feedback; the third and fourth items are to suppress the harmonics of the system, THD u 、THD i are the voltage and current distortion rates, and the fitness function weight coefficients are h1=h2=0.3 and h3=h4=0.2.

[0051] In a second aspect of the present invention, a controller is provided, the controller comprising a processor and a memory storing computer program instructions. When the processor executes the computer program instructions, the harmonic control method of the three-phase SVPWM grid-connected inverter as described in the first aspect is implemented.

[0052] According to a third aspect of the present invention, a computer-readable storage medium is provided, in which at least one executable instruction is stored. When the executable instruction is executed on an electronic device, the electronic device executes the three-phase SVPWM grid-connected inverter harmonic control method as described in the first aspect.

[0053] Beneficial effect: This application proposes a harmonic control method for a three-phase SVPWM grid-connected inverter based on an improved particle swarm algorithm, which is used to optimize the K of the PI controller by using an improved particle swarm optimization (IPSO) algorithm. P and K I The customized improved inertia weight comprehensively considers the minimum, maximum and fitness function of the inertia weight, and the fitness function adds the harmonic suppression performance index. The improved particle swarm optimization algorithm enhances the global search capability and precise search capability, reduces the probability of falling into the local optimal solution, and the optimized K P and K I The parameters are applied to the control system, making the grid-connected process of the grid-connected inverter more stable, enhancing the anti-interference ability, reducing the static error, and improving the power quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is the schematic diagram of the three-phase SVPWM grid-connected inverter.

[0055] Figure 2 Schematic diagram of the voltage-oriented dual closed-loop space vector control model based on dq coordinates.

[0056] Figure 3 This is the structure diagram of the current inner loop transfer function.

[0057] Figure 4 This is the structure diagram of the voltage outer loop transfer function.

[0058] Figure 5 PI parameter structure diagram for improving particle swarm optimization.

[0059] Figure 6 Flowchart for improving particle swarm optimization PI parameters.

[0060] Figure 7 It is a MATLAB simulation model of three-phase grid-connected inverter based on improved particle swarm algorithm.

[0061] Figure 8 It is a MATLAB simulation model of a double closed-loop control system.

[0062] Fig. 9 To improve the particle swarm model simulation model.

[0063] Fig.10 Iterate the change curve to improve the inertia weight.

[0064] Fig.11 It is the iterative change curve of the fitness function.

[0065] Fig.12 It is the PSO-PI parameter optimization curve.

[0066] Fig.13 It is the optimization curve of IPSO-PI parameters.

[0067] Fig.14 is the grid-connected voltage curve.

[0068] Fig.15 is the grid-connected current curve.

[0069] Fig.16 When the disturbance Change curve (PI\IPSO-PI).

[0070] Fig.17 When the disturbance Change curve (PSO-PI\IPSO-PI).

[0071] Fig.18 When the disturbance Local variation curve (PSO-PI\IPSO-PI).

[0072] Fig.19 It is the grid voltage THD and harmonic spectrum.

[0073] Fig. 20 It is the grid current THD and harmonic spectrum. DETAILED DESCRIPTION

[0074] In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is apparent to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features known in the art are not described.

[0075] The following is a detailed technical solution of a three-phase SVPWM grid-connected inverter harmonic control method based on an improved particle swarm algorithm proposed by the present invention, which is described through specific embodiments.

[0076] 1. Construct a three-phase SVPWM grid-connected inverter model

[0077] The main circuit diagram of the three-phase SVPWM grid-connected inverter is as follows: Figure 1 As shown in the figure, the grid-connected inverter part is a three-phase voltage-type full-bridge inverter, and the power electronic switch is composed of 6 insulated gate bipolar transistor (IGBT) modules. C1 is the DC side support capacitor; u a 、u b 、u c is the AC voltage at the output side of the grid-connected inverter; i a 、i b 、i c is the AC current at the output side of the grid-connected inverter; e a 、e b 、e c is the phase voltage of the power grid; u dc The DC voltage is input to the DC side; i dc S is the DC current input to the DC side; n (n=1,2,......,6) is IGBT; R, L and C2 are filter resistance, inductance and capacitance.

[0078] There are 6 power electronic switches, each of which has two states: on and off. The switch function g is defined as x (x=a,b,c), when g x =1, the switches S1, S3, and S5 are turned on. x = 0, the switches S2, S4, and S6 are turned on. The equation in the three-phase stationary coordinate system is:

[0079]

[0080] It needs to be converted into a variable that is easy for PI to track and control. The three-phase stationary coordinate system is converted into a vertical α-β coordinate system, and then the dq transformation is performed. The mathematical model under the dq coordinate is:

[0081]

[0082] Where, -ωLi q +ωLi q is the coupling term, U dc is the DC voltage value, e d and e q is the AC voltage value of the power grid in dq coordinates, i d and i q is the AC current value of the power grid under dq coordinates, R and L are the resistance and inductance values.

[0083] 2. Propose a dual closed-loop space vector control strategy

[0084] The input of the three-phase SVPWM grid-connected inverter is direct current, which is converted into alternating current with the same frequency and phase as the grid voltage through the grid-connected inverter and sent to the grid. The control loop adopts double closed-loop control, with the outer loop being the voltage loop and the inner loop being the current loop. The sampling circuit measures the three-phase output voltage u of the inverter a 、u b 、u c and the current signal i a 、i b 、i c After dq coordinate transformation, the active current i is obtained d and reactive current i q and u d and u q The PI regulator of the voltage outer loop generates Send current to inner loop, with i d 、i q After comparison, the current PI regulator makes i d and i q After decoupling, we get the decoupling control. and After the inverse coordinate transformation, we get and Then, SVPWM is used to control the on and off of the switch tube of the grid-connected inverter to achieve the grid connection of the inverter. The voltage-oriented double closed-loop space vector control model based on dq coordinates is as follows: Figure 2 shown.

[0085] 2.1. Design current inner loop parameters

[0086] The current inner loop can improve the system's influence and reduce the system's steady-state error through PI regulation. The current inner loop is mainly composed of PI link, system delay link, inverter link, filter link, etc. The power electronic switch in the inverter link adopts pulse width modulation technology, which leads to control delay, but the frequency of the modulation wave is much greater than the frequency of the power grid, so this delay is relatively small. At the same time, the system itself also has a certain delay. By decomposing the current inner loop controller, we can get the following: Figure 3 The transfer function structure diagram is shown.

[0087] Figure 3 Medium T s is the sampling period, T PWM is the switching period of the power electronic switch in the grid-connected inverter, K Pd-in , K Id-in , K Pq-in , K Iq-in are the proportional and integral parameters of the PI controller for the d-axis and q-axis of the inner current loop, respectively. The parameters in the LC filter, the transmission line resistance, the coupling inductance value, and the transformer equivalent parameters are comprehensively considered. Through the current inner loop transfer function structure diagram and engineering experience, it can be approximated that the current inner loop is a first-order inertia link, which is convenient for the construction of the voltage outer loop. The estimated current inner loop PI parameter is K Pd-in =K Pq-in =4, K Id-in =K Iq-in =1.05, the system is stable.

[0088] 2.2 Design voltage outer loop parameters

[0089] The voltage outer loop is referenced by the d-axis voltage u dgiven , q-axis voltage reference u qgiven The actual voltage u d 、u q Compare and perform PI control to generate current reference values ​​for the d-axis and q-axis and If it is operated at unity power factor, no q-axis current reference is required, that is, The voltage outer loop can enhance the system's anti-interference capability and improve the stability of the DC bus voltage. Figure 4 The figure shows the structure diagram of the voltage outer loop transfer function.

[0090] Figure 4 Middle K Pd-out , K Id-out , K Pq-out , K Iq-out are the proportional and integral parameters of the PI controller of the voltage outer loop d-axis and q-axis respectively. In this step, the parameters of the LC filter and the equivalent capacitance of the transmission line are comprehensively considered. Similarly, through the voltage outer loop transfer function structure diagram, the closed loop transfer function expression is obtained.

[0091]

[0092] And the approximate value of the voltage outer loop PI parameter is estimated. The estimated voltage outer loop PI parameter is K Pd-out =K Pq-out =1.5, K Id-out =K Iq-out =2, the system is stable.

[0093] 3. Principle of Particle Swarm Optimization Algorithm

[0094] The particle swarm optimization algorithm was proposed in 1995. It simulates the behavior of a flock of birds to find the optimal solution. Each potential solution to the optimization problem is regarded as a particle in the search space. Each particle has its own position and speed. It searches for the optimal solution by constantly updating its speed and position. Each particle update is completed by its own speed vector, individual optimal speed vector and global optimal speed vector. These three speed vectors jointly affect the particle speed update, prompting the particle to continuously find a better solution in the search space.

[0095] Suppose there is an N-dimensional search space R N In the example, the population consists of m particles, and the positions of the particles can be expressed as

[0096] X i ={X i1 ,X i2 ,…X ik ,…X iN} (4)

[0097] Where, X iN Represents the position of the i-th particle in the N-dimensional search space.

[0098] The speed and direction of each particle's flight are determined by V i To decide, V i Expressed as

[0099] V i = {V i1 ,V i2 ,…V ik ,…V iN} (5)

[0100] The particle swarm algorithm first initializes a group of random particles in the N-dimensional search space, and then searches for the optimal solution through iterative updates. ifit and the global extremum G ifit To update its position and flight speed. ifit Refers to the current optimal solution found by the particle itself, G ifit Refers to the optimal solution that the entire population can find. ifit and G ifit Respectively expressed as

[0101] P ifit = {P ifit1 ,P ifit2 ,…P ifitk ,…P ifitN} (6)

[0102] G ifit = {G ifit1 ,G ifit2 ,…Gifitk ,…G ifitN} (7)

[0103] According to the requirements of iterative update, the formula for particle update speed and position is:

[0104]

[0105]

[0106] In the formula, in is a particle in the N-dimensional search space; k is the current iteration number; k+1 is the update iteration number; ω is the inertia weight coefficient, which is a constant and will not be updated; c1 and c2 are acceleration coefficients, representing the particle's acceleration to its own individual extreme value P ifit and the global extremum G ifit The acceleration weight of propulsion; r1 and r2 are random numbers, which are selected between 0 and 1. This number is to maintain the diversity of particles.

[0107] In the particle swarm optimization algorithm, the maximum number of iterations can be set, and the optimal solution can be found after a certain number of iterative update calculations.

[0108] 4. Optimal design of PI parameters of grid-connected inverter based on improved particle swarm algorithm

[0109] The structure diagram of PI parameters of grid-connected inverter optimized by improved particle swarm algorithm is shown in Figure 5 As shown in the figure, the improved particle swarm algorithm is used for the d-axis and q-axis of the voltage outer loop and current inner loop of the grid-connected inverter. First, the approximate range of the PI parameters is determined through the aforementioned engineering calculations, and then the initial values ​​of the inertia weight, search times, population size, initial position, search speed, etc. are set. Through iterative update optimization, the appropriate PI parameters are found, and the PI parameters are replaced by the position of the particle swarm. In the entire optimization process, the fitness function and the improved inertia weight affect the final result of the optimization.

[0110] Particles update their positions and velocities according to formulas (8) and (9), which is a process of positive feedback information, that is, using their own information and their own individual extreme value P ifit and the global extremum G ifit To obtain the position and speed of the next iteration update. ifit >Global extremum G ifit When the inertia weight ω is large, it is easy to enter the local optimal solution, resulting in the failure to search for the global optimal solution, which is not conducive to the global optimization of the system. This problem can be solved by improving the inertia weight. The inertia weight ω has the ability to balance local search and global search. It has the following characteristics: when the inertia weight ω is relatively large, it has a greater guiding significance for global search; on the contrary, when the inertia weight ω is small, local search is prone to occur. When Pifit >G ifit When , the inertia weight can be changed to prevent premature entry into the local optimal solution. The improved inertia weight is as follows:

[0111]

[0112] Where ω max and ω min To improve the minimum and maximum values ​​of the inertia weight ω; ω takes values ​​between 0 and 1, according to experience ω max It can be taken as 0.9, ω min It can be taken as 0.3. k is the current fitness function. min and J pj is the minimum and average value of the fitness function. (k) As the number of iterations and the fitness function change, P ifit >G ifit When , a relatively large inertia weight is given to enhance the global search capability, thereby avoiding falling into the local optimal solution.

[0113] The fitness function plays an important role in PI parameter optimization, and its selection is crucial. The fitness function can be selected as the integration of the square of the error, the integration of the absolute value of the error, or the integration of the product of time and the absolute value of the error. Based on the third method, this paper considers the applicability of the project and adds the THD suppression index. Its fitness function is

[0114] J k =h1∫0 M t(|e ud (t)|+|e uq (t)|)dt+h2∫0 M t(|e id (t)|+|e iq (t)|)dt+h3∫0 M THD u dt+h4∫0 M THD i dt (11)

[0116] The first and second terms (integral terms) in the formula are to ensure the response speed and stability of the system, M is the data length, e ud (t), e uq (t), e id (t), e iq (t) is the difference between the given voltage and current of the d-axis and q-axis and the actual voltage and current of the feedback; the third and fourth items are to suppress the harmonics of the system, THD u 、THD iare the voltage and current distortion rates, and the fitness function weight coefficients are h1=h2=0.3 and h3=h4=0.2.

[0117] After clarifying the improved inertia weight and fitness function, the grid-connected inverter PI parameter optimization can be carried out step by step. The improved particle swarm optimization PI parameter flow chart is as follows Figure 6 shown.

[0118] The specific steps are described as follows:

[0119] (1) Determine the initialization parameters. The population size SIZE is 20. The search space dimension N is 8 dimensions, representing K Pd-in , K Id-in , K Pq-in , K Iq-in , K Pd-out , K Id- , K Pq-ou , K Iq-out The maximum number of iterations n is 50. The acceleration coefficient c1 is 1.5 and c2 is 2. The search range of PI parameters is between 0 and 50. The fitness function weight coefficients h1=h2=0.3 and h3=h4=0.2. The maximum inertia weight ω max is 0.9, the minimum inertia weight ω min is 0.3. The random numbers r1 and r2 are generated by the random function rand(*). The initial position and velocity of the particle are random values ​​and zero values ​​within the PI parameter search range.

[0120]

[0121]

[0122] (2) Build the fitness function of formula (11) in the SIMULINK model, and use the Sphere function to establish a connection with the m file to calculate the fitness value and determine the optimal value of the initial individual and the optimal value of the initial population.

[0123] (3) Update the particle speed and position according to formulas (8) and (9), obtain a new population, and calculate the current fitness function value.

[0124] (4) Compare the individual optimal value of the particle with the current fitness function value. If the current fitness function value is less than the individual optimal value, the current fitness function value is the current optimal value.

[0125] (5) Compare the optimal value of the population with the current fitness function value. If the current fitness function value is less than the optimal value of the population, the current fitness function value is the current optimal value.

[0126] (6) Determine the index and range requirements. When the harmonic suppression performance index is met, or the parameter setting range is exceeded or the number of iterations is reached, the iteration update is terminated and the optimal value G is given. fit (K Pd-in , K Id-in , K Pq-in , K Iq-in , K Pd-ou , K Id-out , K Pq-out , K Iq-out ). Otherwise, return to step (2) and re-enter the loop.

[0127] 5. Algorithm simulation

[0128] In order to study the PI controller parameters, a three-phase grid-connected inverter model based on the improved particle swarm algorithm was built in MATLAB, and the SVPWM control mode was used. The improved particle swarm algorithm was used for the PI parameters of the voltage outer loop and the current inner loop, and the rationality of the parameter optimization was verified through simulation analysis. The simulation parameters are shown in Table 1.

[0129] Table 1 System simulation parameters

[0130] parameter Numeric unit Grid-connected phase voltage 220 V DC bus voltage 800 V Switching frequency 20 kHz AC side inductor equivalent resistance 0.1 Ω AC side inductance 5 mh AC side capacitor 3 μF

[0131] The MATLAB simulation model of the three-phase grid-connected inverter based on the improved particle swarm algorithm is as follows Figure 7 As shown. The model includes a DC power supply, a three-phase bridge converter, an LC filter, a grid-connected load, an improved particle swarm model, an SVPWM model, and a control model. The control model also includes park transformation, phase-locked loop control, decoupling control, voltage outer loop control, and current inner loop control. The phase-locked loop control link can lock the grid voltage to ensure that the grid voltage and current are in phase. The MATLAB simulation model of the dual closed-loop control part is shown in Figure 8 The improved particle swarm simulation model is shown in Fig. 9 shown.

[0132] The model is run using the PI algorithm, PSO-PI algorithm, and IPSO-PI algorithm, and the corresponding parameter values ​​are shown in Table 2.

[0133] Table 2 PI control parameter values ​​of grid-connected inverter

[0134]

[0135] Fig.10 In order to improve the iterative change curve of the inertia weight, it can be seen that the value of the inertia weight changes nonlinearly and stops changing after 29 iterations. The improved inertia weight makes it less likely for the optimization to fall into the local optimal solution and improves the optimization convergence speed. Fig.11The iterative change curves of the fitness function of PSO and IPSO. Both sets of curves have a decreasing historical optimal value as the iterations continue to increase, indicating that the optimization process is constantly approaching the global optimal value with each iterative update of the particles. When the search meets the set value range and performance indicators and finds the optimal position, the iterative update is stopped. It can be seen that PSO and IPSO have basically found their respective optimal values ​​after about 30 iterations, and the IPSO optimization process converges slightly faster than PSO. Fig.12 is the PI parameter optimization curve of PSO, Fig.13 This is the PI parameter optimization curve of IPSO. It can be seen from it that the drastic change of parameters is basically completed before 30 times. It can be approximately considered that the optimal solution of PI parameters is basically found, and the particles basically stop iterating and updating.

[0136] Fig.14 The three-phase voltage curve after the inverter is connected to the grid is given. It can be seen that the PI algorithm can allow the system to quickly enter a stable grid-connected state, but the rise time is the slowest in the dynamic situation, and the harmonics are larger than those under the IPSO-PI algorithm at the peak of the grid. The system rise time under the PSO-PI algorithm is the fastest, but the harmonic content is too much, and the voltage continues to fluctuate. The system rise time and overshoot under the IPSO-PI algorithm are moderate, 5ms and 0.92% respectively. In addition, the overall harmonics are minimal, and the overall performance is moderate. Fig.15 The three-phase current curve after the inverter is connected to the grid is given. The overall performance is consistent with the voltage situation and will not be repeated here.

[0137] Fig.16 , Fig.17 and Fig.18 The system is given by Change curve. At 0.7s, the system adds a step disturbance, from Fig.16 As can be seen from the figure, the impact of PI is much higher than that of IPSO-PI. Fig.17 It can be seen that the impact of PSO-PI is slightly higher than that of IPSO-PI. Fig.18 It can be seen that when the system is stable, under the IPSO-PI algorithm, At -0.25 Small-scale adjustments are performed nearby; while under the PSO-PI algorithm, In 1 The dynamic performance of IPSO-PI is better than that of PSO-PI and even better than that of PI.

[0138] When the grid-connected inverter is connected to the grid, high-quality grid-connected voltage and current are required to provide better power to users. Reducing the harmonics of voltage and current and maintaining a low THD are one of the indicators to ensure high-quality electric energy. IEEE Std.929-2000 requires that the THD of grid-connected AC voltage and current be less than 5%.

[0139] Fig.19 And Fig. 20 The sample mean THD and harmonic spectrum of grid-connected voltage and current are given respectively. It can be seen from the figure that the fundamental wave content of 50Hz is very high, the harmonics of each order are relatively low, and the distribution is also roughly the same, but the harmonic content has some differences. The method for extracting the sample mean THD is as follows: In the "FFT Analysis" of "POWERGUI", select the waveform of the stable output part for display, and set the parameters "Number of cycles" to 20, "Fundamental frequency" to 50, and "Max frequency" to 1000.

[0140] Fig.19 (a) adopts the PI control technology, Fig.19 (b) optimizes the PI control with the particle swarm algorithm, Fig.19 (c) adopts the improved particle swarm algorithm to optimize the PI control. It can be seen that for the sample mean THD, IPSO-PI < PSO-PI < PI, and the voltage harmonic content is reduced by 2.84 percentage points. For the grid-connected current, Fig. 20 (a) is the PI algorithm, Fig. 20 (b) is the PSO-PI algorithm, Fig. 20 (c) is the IPSO-PI algorithm. The sample mean THD of the grid-connected current is reduced from 2.97% to 1.72%, and finally to 0.32%. The current harmonic content is reduced by 2.65 percentage points. The improved particle swarm optimization algorithm is effective in suppressing voltage and current harmonics.

[0141] As a preferred embodiment, the present application also discloses a controller, which includes a processor and a memory storing computer program instructions. When the processor executes the computer program instructions, the above-mentioned harmonic control method of the three-phase SVPWM grid-connected inverter is realized. The specific content will not be elaborated here.

[0142] As a preferred embodiment, the present application also discloses a computer-readable storage medium, which stores at least one executable instruction. When the executable instruction runs on an electronic device, the electronic device executes the harmonic control method of the three-phase SVPWM grid-connected inverter as described in the first aspect. The specific content will not be elaborated here.

[0143] In summary, the present application discloses a modeling, simulation and testing method for a high-performance control strategy based on a three-phase SVPWM grid-connected inverter. Simulation and verification were performed in MATLAB, and operation was performed in a physical platform with good operation results. The advantage of the IPSO algorithm is that it minimizes the errors of the voltage outer loop controller and the current inner loop controller, thereby obtaining the optimal parameters of the PI controller, and at the same time, a harmonic suppression parameter is added to the objective function to achieve the optimal harmonic suppression effect.

[0144] This application first proposes the main circuit topology of the three-phase SVPWM grid-connected inverter, determines the dual closed-loop control strategy based on space vector modulation according to the mathematical model, and proposes the IPSO algorithm based on the PSO algorithm. The IPSO algorithm proposed in this article optimizes the inertia weight and fitness function, enhances the global search capability and precise search capability, and reduces the probability of falling into the local optimal solution. SIMULINK simulation was performed on the PI algorithm, PSO-PI algorithm and IPSO-PI algorithm. From the simulation results, the following conclusions can be drawn about the IPSO-PI algorithm: the algorithm effectively suppresses harmonics, and the sample average THD of the grid-connected voltage and grid-connected current has dropped by 2.84 and 2.56 percentage points. In addition, the algorithm enables the system to have good anti-interference ability. When the system is switched and disturbed by load, the reaction speed and the suppression of shock are higher than the PSO-PI algorithm, and it has a smaller static error. In steady state There is a fluctuation of about 0.5, the system rise time and overshoot are 0.5ms and 0.92% respectively, and the overall system response is improved. In short, this study provides a solution and idea for optimizing PI controller parameters and improving system grid-connected performance.

[0145] As described above, although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the present invention itself. Various changes in form and details may be made without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A harmonic control method for a three-phase SVPWM grid-connected inverter based on an improved particle swarm algorithm, characterized in that: The steps include: S1. Construct a main circuit model of a three-phase SVPWM grid-connected inverter, convert direct current into alternating current with the same frequency and phase as the grid voltage through the three-phase SVPWM grid-connected inverter and send it into the grid; the control loop of the three-phase SVPWM grid-connected inverter adopts double closed-loop control, including a voltage outer loop and a current inner loop; S2. Determine the range of voltage outer loop parameters and current inner loop parameters through engineering calculations; S3, using the improved particle swarm algorithm to find the optimal parameters within the range of the voltage outer loop parameters and the current inner loop parameters, and outputting the improved particle swarm optimization PI parameters; S4. Use the improved particle swarm optimization PI parameters to complete the harmonic control of the three-phase SVPWM grid-connected inverter.

2. The three-phase SVPWM grid-connected inverter harmonic control method according to claim 1, characterized in that: The main circuit model of the three-phase SVPWM grid-connected inverter includes a three-phase voltage-type full-bridge inverter and six power electronic switches; The power electronic switch is composed of an insulated gate bipolar transistor; each power electronic switch has two states: on and off; a switch function g is defined x , x=a,b,c; when g x =1, the switches S1, S3, and S5 are turned on; when g x =0, the switches S2, S4, and S6 are turned on; The equation in the three-phase stationary coordinate system is: Where U dc The DC voltage is input to the DC side; i x is the AC current at the output side of the grid-connected inverter; e x is the phase voltage of the power grid; R is the filter resistance value; L is the inductance value.

3. The harmonic control method of the three-phase SVPWM grid-connected inverter according to claim 2, characterized in that: Determining the range of the voltage outer loop parameters and the current inner loop parameters by engineering calculation specifically includes: The three-phase stationary coordinate system is converted into a vertical α-β coordinate system, and then a dq transformation is performed. The mathematical model under the dq coordinate is: Where, -ωLi q +ωLi q is the coupling term; U dc is the DC voltage value; i d and i q are the AC current values ​​of the power grid in dq coordinates respectively; e d and e q are the AC voltage values ​​of the power grid in dq coordinates respectively; g d and g q are the switching functions decomposed into the d-axis and q-axis respectively; The current inner loop is composed of PI controller link, system delay link, inverter link, and filter link; The voltage outer loop is referenced by the d-axis voltage u dgiven , q-axis voltage reference u qgiven The actual voltage u d 、u q Compared with PI control, the closed-loop transfer function expression is obtained: Where G1(s) is the transfer function of the voltage outer loop PI controller; G2(s) is the transfer function of the current inner loop; G4(s) is the feedback sampling delay; in: Where, T S is the sampling period; K Pd- , K Id-in , K Pq-in , K Iq-i are the proportional and integral parameters of the PI controller of the d-axis and q-axis of the inner current loop respectively; K Pd-out , K Id-ou , K Pq-out , K Iq-out are the proportional and integral parameters of the PI controller of the d-axis and q-axis of the voltage outer loop respectively; Above K Pd-in , K Id-in , K Pq-in , K Iq-in , K Pd-out , K Id-o , K Pq-out , K Iq-out Together they constitute the PI parameters, and the range of the PI parameters is preliminarily screened out.

4. The harmonic control method of the three-phase SVPWM grid-connected inverter according to claim 3, characterized in that: The method of finding the optimal parameters by using the improved particle swarm algorithm and outputting the improved particle swarm optimization PI parameters specifically includes: Set the particle position, particle flight speed, number of iterations, and inertia weight, and find the PI parameters that meet the predetermined conditions through iterative update optimization; the PI parameters are replaced by the position of the particle group; During the entire optimization process, the fitness function and the improved inertia weight affect the final result of the optimization.

5. The harmonic control method of the three-phase SVPWM grid-connected inverter according to claim 4, characterized in that: An N-dimensional search space R N In the example, the population consists of m particles, and the particle positions are expressed as: X i ={X i1 ,X i2 ,…X ik ,…X iN } Where, X iN Represents the position of the i-th particle in the N-dimensional search space; The particle flight speed is expressed as: V i ={V i1 ,V i2 ,…V ik ,…V iN } Initialize a group of random particles in the N-dimensional search space, and then find the optimal solution through iterative updates. In each iterative update, search for its own individual extreme value P ifit and the global extremum G ifit To update its position and flight speed; Where P ifit Refers to the current optimal solution found by the particle itself, G ifit Refers to the optimal solution that the entire population can find; P ifit and G ifit Respectively expressed as: P ifit ={P ifit1 ,P ifit2 ,…P iftk ,…P iftN } G ifit ={G ifit1 ,G ifit2 ,…G ifitk …G ifitN } According to the requirements of iterative update, the formula for particle update speed and position is: In the formula, in is a particle in the N-dimensional search space; k is the current iteration number; k+1 is the update iteration number; ω is the inertia weight coefficient; c1 and c2 are acceleration coefficients, representing the particle's movement toward its own individual extreme value P ifit and the global extremum G ifit The acceleration weight of the propulsion; r1 and r2 are random numbers selected between 0 and 1.

6. The harmonic control method of the three-phase SVPWM grid-connected inverter according to claim 4, characterized in that: The expression of the inertia weight is as follows: In the formula, ω max and ω min To improve the minimum and maximum values ​​of the inertia weight ω, it takes values ​​between 0 and 1; J k is the current fitness function; J min and J pj is the minimum and average value of the fitness function; ω (k) As the number of iterations and the fitness function change, P ifit >G ifit , give greater inertia weight.

7. The harmonic control method of the three-phase SVPWM grid-connected inverter according to claim 6, characterized in that: The fitness function J k The expression is as follows: In the formula, the first and second integral terms are to ensure the response speed and stability of the system; M is the data length; e ud (t), e uq (t), e id (t), e iq (t) is the difference between the given voltage and current of the d-axis and q-axis and the actual voltage and current of the feedback; the third and fourth items are to suppress the harmonics of the system, THD u 、THD i are the voltage and current distortion rates, and the fitness function weight coefficients are h1=h2=0.3 and h3=h4=0.

2.

8. A controller, characterized in that: The controller comprises: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the harmonic control method of the three-phase SVPWM grid-connected inverter according to any one of claims 1 to 7 is implemented.

9. A computer-readable storage medium, characterized in that: The storage medium stores at least one executable instruction. When the executable instruction is executed on the electronic device, the electronic device executes the three-phase SVPWM grid-connected inverter harmonic control method according to any one of claims 1 to 7.

Images