Sliding mode controller and phase-locked loop parameter multi-objective optimization method based on grid-connected inverter
The multi-objective optimization of grid-connected inverter parameters using an improved NSGA-II algorithm and a sliding mode controller addresses the complexity of competing objectives, enhancing dynamic performance and stability, ensuring robust operation under varying grid conditions.
Patent Information
- Application Number
- CN202411206406.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-03
AI Technical Summary
Traditional methods for optimizing grid-connected inverter parameters often rely on single-objective strategies, failing to address the complexity of multiple competing objectives, which affects the dynamic performance and stability of the inverter under varying power grid conditions.
A multi-objective optimization method using an improved NSGA-II algorithm, combined with a sliding mode controller and phase-locked loop, optimizes parameters to enhance the dynamic performance and stability of grid-connected inverters by synchronizing the output current with the grid voltage, reducing harmonic content, and improving robustness against grid disturbances.
The method significantly enhances the waveform quality of inverter output current, improves dynamic response speed, and maintains system stability under grid fluctuations, reducing the risk of disconnection and hardware costs while supporting the widespread application of grid-connected inverters in renewable energy systems.
Smart Images

Figure CN119070382A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of power grid control technology, and in particular to a sliding mode controller based on a grid-connected inverter and a multi-objective optimization method for phase-locked loop parameters. Background Art
[0002] The grid-connected inverter is an important power electronic device whose main function is to convert direct current into alternating current and realize parallel operation with the public power grid. With the transformation of the global energy structure and the rapid development of clean energy, the grid-connected inverter plays an increasingly important role in photovoltaic power generation, wind power generation and other fields. In the design and control of the grid-connected inverter, sliding mode control and phase-locked loop are two important technical means, which are used to improve the dynamic performance of the system and the ability to track the frequency and phase of the power grid respectively.
[0003] The sliding mode controller can achieve fast tracking and anti-interference capabilities of the output current of the grid-connected inverter, ensuring that the output current can quickly and accurately track the reference current. It can effectively overcome the influence of factors such as inverter parameter changes and external interference, ensuring that the system still has good performance under uncertain conditions. The phase-locked loop is a phase tracking technology that can lock the frequency and phase of the power grid to ensure that the inverter output is synchronized with the power grid. Sliding mode control and phase-locked loop can be combined with each other to form an efficient and robust control strategy. Through this combination, the grid-connected inverter can maintain good performance under different power grid conditions, ensuring the stable operation of the new energy power generation system and the improvement of power quality.
[0004] Traditional methods often rely on experience or single-objective optimization strategies, such as cost- or efficiency-based optimization. However, these methods may not be able to handle the complexity of multiple competing objectives. NSGA-II generates a series of non-dominated solutions by continuously evolving and evolving individuals in the population. These solutions constitute a solution set that approximates the Pareto frontier. Engineers can select the optimal parameter combination from this solution set and make further adjustments and optimizations based on specific project requirements.
[0005] In summary, the NSGA-II algorithm provides an efficient and comprehensive optimization strategy for optimizing grid-connected inverter parameters, which can find the best balance between multiple competing objectives, thereby improving the overall performance and economy of the solar power generation system. Summary of the invention
[0006] Purpose of the invention: The present invention proposes a multi-objective optimization method for sliding mode controller and phase-locked loop parameters based on a grid-connected inverter, which can simultaneously optimize the sliding mode controller and the phase-locked loop parameters, significantly improving the dynamic performance of the grid-connected inverter and the grid-connected current quality.
[0007] Technical solution: The present invention provides a multi-objective optimization method for a sliding mode controller and a phase-locked loop parameter based on a grid-connected inverter, comprising the following steps:
[0008] (1) Based on the filter inductor current and filter capacitor voltage of the three-phase grid-connected inverter, an equivalent system model of the three-phase grid-connected inverter in the α-β coordinate system is constructed;
[0009] (2) Based on the equivalent system model, the sliding surface is designed and the sliding mode controller of the three-phase inverter is designed;
[0010] (3) Design a phase-locked loop to ensure that the given phase of the grid-side current matches the phase of the grid voltage, so that the output current frequency and phase are synchronized with the grid voltage;
[0011] (4) Propose an improved NSGAII algorithm to optimize the determined objective function; the improved NSGAII algorithm improves the diversity and convergence of the NSGAII algorithm by using a piecewise chaotic mapping initialization method and a linearly decreasing adaptive mutation operator and strengthening the dominance relationship;
[0012] (5) The optimal parameters of the phase-locked loop and the sliding mode controller are optimized through the improved NSGAII algorithm, and the optimal parameters are sent to the phase-locked loop and the sliding mode controller. The phase-locked loop is phase-locked through the integral link, and the sliding mode controller outputs the control rate;
[0013] (6) According to the output control rate obtained in step (5), the full-bridge switch of the three-phase grid-connected inverter is controlled by SPWM to achieve phase locking and current control.
[0014] Furthermore, the equivalent system model on the α-β coordinate system in step (1) is:
[0015]
[0016] in, and They represent the current in the AC side inductance of the inverter on the α-axis and β-axis respectively. and Respectively represent the voltage of the AC side capacitor of the inverter on the α-axis and β-axis, u conα and u conβ K is the control voltage output by the inverter on the α-axis and β-axis; PWM is the inverter equivalent gain, which is used to convert the control input signal into the actual switching signal.
[0017] Furthermore, the sliding mode controller of the three-phase inverter in step (2) is:
[0018]
[0019] Among them, u con is the control rate, i.e. the modulation signal of the inverter, which is used to control the switching action of the inverter and thus affect the output current; c 1 is a positive coefficient of sliding mode control, which is used to adjust the dynamic performance of the equivalent system; K is used to ensure the arrival condition of the sliding surface, that is, the system state can reach and remain on the sliding surface; η is used to adjust the influence of the sign function to reduce chattering; sign(s) is a sign function, which is used to generate nonlinear control action to ensure that the system state slides on the sliding surface.
[0020] Furthermore, the phase-locked loop in step (3) includes a phase detector, a loop filter, a voltage-controlled oscillator and a frequency divider; the phase detector compares the phase difference between the grid voltage and the internal synchronization signal of the control system, and converts the phase difference into a voltage signal; the voltage signal is processed by the loop filter and used to control the voltage-controlled oscillator, thereby adjusting the frequency and phase of the internal synchronization signal of the system to keep it synchronized with the grid voltage.
[0021] Furthermore, the improved NSGAII algorithm in step (4) is:
[0022] The initialization formula of Piecewise chaotic mapping is:
[0023]
[0024] Among them, P is a piecewise control factor used to divide the piecewise function, x i It is the state value generated by iteration, and its size is between 0 and 1. Due to its piecewise linear characteristics, PWLCM can theoretically provide better randomness and complexity, and can construct an initial population with a wide range, uniform distribution, and strong global search capabilities.
[0025] The enhanced dominance relation SDR is used to solve the convergence and diversity of each iterative solution set; when a candidate solution x dominates another candidate solution y if and only if:
[0026]
[0027] in, Used to measure the degree of convergence of x, θ xy represents the acute angle between the target values of two candidate solutions, i.e., θ xy =arccos(f(x),f(y)); It is the size of the specific region of each candidate solution in the solution space;
[0028] Adopt the linearly decreasing adaptive mutation probability operator strategy LDM to dynamically adjust the mutation probability:
[0029] MP=MPmax -(MP max -MP min )*(FE max -FE) / FE max
[0030] In the formula, MP max is the maximum mutation probability, MP min is the minimum mutation probability, FE max is the maximum number of iterations, and FE is the current number of iterations.
[0031] Furthermore, the objective function in step (4) is:
[0032]
[0033] Objective function f 1 The phase difference between the reference voltage and the actual current is the smallest, that is, the phase-locked loop achieves phase locking when the input and output phase angles are equal. Calculate Δφ=φv[X k ]-φi[X k ], X k is the frequency domain signal; the objective function f 2 The difference between the reference current and the actual current is the smallest; i ref is the reference current i l is the actual current.
[0034] Furthermore, the implementation process of step (5) is as follows:
[0035] (51) Set the parameter range, where the phase-locked loop optimization parameter is c 1 and k, the optimized parameters of the phase-locked loop are K P and K I ; and initialize the population to generate a population P containing N random solutions n , and for the population P n The solution in is normalized; the obtained parameters are brought into the established model, and the fitness of the population is calculated according to the objective function established previously;
[0036] (52) Using the SDR method, the population P n Perform non-dominated sorting to form multiple fronts and calculate the crowding distance of each front; enter the main loop to check whether the termination criteria are met. If not, continue to execute (53);
[0037] (53) According to the position and crowding distance of each solution in the non-dominated frontier, N parents are selected to form a new population P. n ; and for P n Perform mutation operation to generate a new population P n ; Calculate the new population P nThe fitness of the population P is n Perform non-dominated sorting to form multiple fronts, calculate the crowding distance of each front, delete redundant solutions from the front with the worst crowding distance, and update the population;
[0038] (54) A set of Pareto optimal solutions is obtained, and a set of parameters closest to the coordinate origin is selected from the Pareto optimal solution set to achieve that the output current phase of the output inverter is equal to the grid voltage phase, and the sliding mode controller makes the output current equal to the reference current.
[0039] Furthermore, the implementation process of step (6) is as follows:
[0040] Firstly, the output of the three-phase inverter is transformed from the abc coordinate to the α-β coordinate by Clarke transformation, and the extended state observer model is established through the error between the output current and the reference current; secondly, the parameters optimized by the improved NSGAII algorithm are introduced into the phase-locked loop and the sliding mode controller. The phase-locked loop adjusts the frequency and phase of its output signal so that it gradually synchronizes with the grid voltage signal and the phase difference remains constant; the sliding mode controller obtains the total sliding mode control rate u by introducing the error e between the output current and the reference current and the sliding surface s into the sliding mode controller. conα and u con β; Finally, u conα and u con The two control rates β are input into the SPWM modulation strategy to generate six PWM signals to control the switch control signals of the six switching devices, so that the DC output current tracks the given value.
[0041] Beneficial effects: Compared with the prior art, the present invention has the following beneficial effects:
[0042] 1. The sliding mode controller adopted in the present invention has optimized sliding mode controller parameters, which significantly improves the waveform quality of the inverter output current and reduces the harmonic content, effectively enhancing the robustness of the system to grid disturbances and ensuring the stable operation of the inverter in a complex grid environment;
[0043] 2. The optimized phase-locked loop parameters adopted by the present invention enable the inverter to track the frequency and phase of the grid voltage more accurately. Even in the case of grid voltage fluctuations or nonlinear load changes, the system can maintain good stability, reducing the risk of inverter disconnection due to grid disturbances;
[0044] 3. The present invention utilizes the application of the improved NSGA-II algorithm to achieve the optimal configuration of control parameters, further improving the dynamic response speed and steady-state performance of the inverter; through parameter optimization, the system reduces the control complexity and hardware cost while maintaining high-performance control effects, providing strong support for the widespread application of grid-connected inverters in the field of renewable energy power generation. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Figure 1 It is a schematic diagram of the structure of a three-phase grid-connected inverter;
[0046] Figure 2 It is a simplified model diagram of an independent single-phase full-bridge inverter circuit;
[0047] Figure 3 It is the structural diagram of the phase-locked loop;
[0048] Figure 4 It is a diagram of the optimization steps of the improved NSGAII algorithm of the present invention.
[0049] Figure 5 It is the control strategy system framework diagram of the present invention;
[0050] Figure 6 The output current and grid voltage waveforms and tracking effect diagram of phase A when the controller of the present invention is started under rated working conditions. DETAILED DESCRIPTION
[0051] The present invention will be further described in detail below in conjunction with the accompanying drawings.
[0052] The present invention discloses a sliding mode controller and a phase-locked loop parameter multi-objective optimization method based on a grid-connected inverter, which mainly includes the following steps:
[0053] Step 1: According to the filter inductor current and filter capacitor voltage of the three-phase grid-connected inverter, an equivalent system model of the three-phase grid-connected inverter on the α-β coordinate system is constructed.
[0054] Figure 1 The schematic diagram of the three-phase grid-connected inverter used in the present invention is composed of a DC voltage source, six IGBT power switches, resistors, filter inductors and a power grid. dc is a DC voltage source, L is a filter inductor, r is a resistor, i a 、i b 、i c Represents the three-phase currents of a, b, and c respectively, and u as 、u bs 、u cs Represents the three-phase grid voltages a, b, and c respectively, and u ak 、u bk 、u ckRepresent the output voltages of the three-phase inverter a, b, and c respectively.
[0055] Since the three-phase inverter is decomposed into three completely independent control systems under the abc stationary coordinates, there is no coupling correlation problem. In terms of control strategy, the three-phase inverter can also be controlled independently under the three-phase stationary coordinates, so the topological structure of the three-phase inverter is equivalently decomposed into three single-phase full-bridge inverter circuits, such as Figure 2 shown.
[0056] According to the equivalent model, the current equation in the abc coordinate system can be written:
[0057]
[0058] Take the inductor current As the state parameter, we can get:
[0059]
[0060] In the formula, u k 、u s are the voltage output on the AC side and the capacitor voltage respectively; i l is the current on the AC side inductor.
[0061] The formula for isopower transformation from abc coordinates to α-β coordinates is:
[0062]
[0063] make
[0064] The mathematical expression of formula (2) can be transformed into the α-β coordinate system by formula (3):
[0065]
[0066] The new state equations are obtained as follows:
[0067]
[0068] in, and They represent the current in the AC side inductance of the inverter on the α-axis and β-axis respectively. and Respectively represent the voltage of the AC side capacitor of the inverter on the α-axis and β-axis, u conα and u conβ It is the control voltage output by the inverter on the AC side on the α-axis and β-axis. k =D i V dc =K PWM u conk , where Di =u con / V tri Indicates the duty cycle within a switching cycle; K PWM The inverter PWM modulator is K PWM =V dc / V tri , used to convert the control input signal into the actual switching signal, u con is the amplitude of the low-frequency sine wave signal, V tri is the amplitude of the high frequency triangular carrier signal.
[0069] Step 2: Based on the equivalent system model, design the sliding surface and the sliding mode controller of the three-phase inverter.
[0070] The current error e is defined as:
[0071] e=i ref -i l (6)
[0072] Define the sliding surface s:
[0073]
[0074] Take the Lyapunov function:
[0075]
[0076] Then its derivative is:
[0077]
[0078] In order to make the derivative of the Lie function negative definite, an exponential approaching sliding mode control rate can be designed:
[0079]
[0080] Among them, u con is the control rate, i.e. the modulation signal of the inverter, which is used to control the switching action of the inverter and thus affect the output current; i ref is the reference current value; c 1 is a positive coefficient in the sliding mode control design, which is used to adjust the dynamic performance of the system. k is a positive coefficient in the sliding mode control design, which is used to ensure the arrival condition of the sliding surface, that is, the system state can reach and remain on the sliding surface; η is a positive coefficient in the sliding mode control design, which is used to adjust the influence of the sign function to reduce chattering; sign(s) is a sign function, which is used to generate nonlinear control action to ensure that the system state slides on the sliding surface. The sliding surface designed by the present invention can maintain the stability and rapidity of the system under the conditions of grid frequency fluctuations and load mutations, and realizes high-quality control of the inverter output current.
[0081] Step 3: Design a phase-locked loop to ensure that the given phase of the grid-side current matches the phase of the grid voltage, so that the output current frequency and phase are synchronized with the grid voltage.
[0082] The parameters of the phase-locked loop have a significant impact on the stability of the system. By optimizing these parameters, the stability problems that may be caused by the phase-locked loop can be effectively solved. In order to achieve unity power factor operation of the PWM rectifier control system, it is necessary to ensure that the given phase of the grid-side current matches the phase of the grid voltage. This requires the system to accurately and in real time detect the phase information of the grid voltage, so phase tracking becomes a key link in the system.
[0083] The phase-locked loop is mainly composed of a phase detector, a loop filter, a voltage-controlled oscillator and a frequency divider. Its structure is as follows: Figure 3 As shown, where θ and ω are the estimated phase and estimated frequency, respectively, K P and K I are the parameters of the PLL proportional integral controller. Its working principle is that the phase detector compares the phase difference between the grid voltage and the internal synchronization signal of the control system, and converts this phase difference into a voltage signal. Subsequently, the voltage signal is processed by the loop filter and used to control the voltage-controlled oscillator, thereby adjusting the frequency and phase of the internal synchronization signal of the system to keep it synchronized with the grid voltage.
[0084] Step 4: Parameter tuning based on the improved NSGAII algorithm. The improved NSGAII algorithm uses the Piecewise chaotic mapping initialization method and the linearly decreasing adaptive mutation operator to enhance the dominance relationship and improve the diversity and convergence of the NSGAII algorithm.
[0085] NSGA-Ⅱ is a multi-objective optimization algorithm that uses fast non-dominated sorting based on genetic algorithms. It is an optimization algorithm based on dominance relations and is widely used in complex multi-objective optimization problems. Its core lies in the stratification based on the dominance relations between individuals before the selection operation is executed. The sliding mode controller and phase-locked loop parameters are adjusted by using its advantages such as not being easily trapped in the local optimum, being easier to find the global optimal solution, and running faster.
[0086] Before the selection operation, the NSGA-II algorithm sorts the population according to the dominance and non-domination relationship between individuals. First, all non-dominated individuals are identified, which are neither dominated by other individuals, forming the first non-dominated optimal layer, and assigning them a shared virtual fitness value. Next, the stratified individuals are ignored, and the remaining individuals continue to be stratified according to the dominance and non-domination relationship, and are assigned a lower new virtual fitness value to ensure that it is smaller than the value of the previous layer. This process continues until all individuals in the population are divided into different layers. Individuals in each layer have the same non-dominated order, and different layers are sorted according to the size of the non-dominated order.
[0087] Since the NSGAII algorithm often falls into a local extreme value before obtaining the global optimal solution when solving complex nonlinear problems, premature convergence occurs. Therefore, in order to avoid the algorithm from falling into a local extreme value at an early stage, the present invention proposes an improved NSGAII algorithm, such as Figure 4 As shown, by utilizing the reinforced dominance relationship, using the Piecewise chaotic map initialization, and introducing the linearly decreasing adaptive mutation probability operator into the mutation operator to improve the diversity of the population, the diversity and convergence of the algorithm are improved.
[0088] Piecewise chaotic mapping has good statistical performance and is a piecewise linear chaotic mapping. It is a chaotic system that divides the domain of chaotic mapping into several intervals and uses different linear functions in each interval to define the mapping relationship. Piecewise is characterized by simple structure, easy implementation, and good chaotic characteristics under certain parameter configurations.
[0089] The chaotic mapping formula is:
[0090]
[0091] Where P is a piecewise control factor used to divide the piecewise function. i The size of is between 0 and 1. Due to its piecewise linear characteristics, PWLCM can theoretically provide better randomness and complexity, and can construct an initial population with a wide range, uniform distribution, and strong global search capabilities.
[0092] The proposed Strengthened Dominance Relation (SDR) solution set has convergence and diversity. A candidate solution x dominates another candidate solution y if and only if
[0093]
[0094] in, Con(x) is used to measure the degree of convergence of x, θxy represents the acute angle between the target values of two candidate solutions, i.e., θ xy =arccos(f(x),f(y)). is the size of a specific region in the solution space for each candidate solution. In order to better handle optimization problems where the target scale is poor or the ideal point is not the origin, when calculating Con(x) and θ xy Previously, the objective values of all candidate solutions should be normalized according to the ideal point and the lowest point of the Pareto front. However, since the Pareto front is not known a priori, the minimum value of each objective and the maximum value of each objective in the population are taken as the ideal point and the lowest point, respectively.
[0095] This enhanced dominance relationship based on the angles between candidate solutions allows the best converged candidate solution to be identified as a non-dominated solution, balancing convergence and maintaining solution diversity during the optimization process.
[0096] The mutation operator is an important part of the genetic algorithm. If the mutation probability is too small, it is easy to cause the population to fall into the local optimum. If the probability is too large, it may cause the population to converge slowly or even not converge. On this basis, the linear decreasing adaptive mutation probability operator (LinearDecreasing Mutation, LDM) strategy is adopted to dynamically adjust the mutation probability:
[0097] MP=MP max -(MP max -MP min )*(FE max -FE) / FE max (13)
[0098] In the formula, MP max is the maximum mutation probability, MP min is the minimum mutation probability, FE max is the maximum number of iterations, and FE is the current number of iterations. In the early stage of iteration, the mutation rate is high, and widely distributed solutions can be generated, which can improve the diversity of the algorithm; in the middle of the iteration, the mutation probability will gradually decrease; in the late stage of iteration, the mutation probability is the smallest, and the solutions obtained are more dense, which can improve the convergence of the algorithm.
[0099] Objective function f 1 The phase difference between the reference voltage and the actual current is the smallest: For the time domain signal, it can be decomposed into a series of sinusoidal signals in the frequency domain, each sinusoidal signal has its own amplitude and phase.
[0100] Fast Fourier transform (FFT) combined with phase analysis was used to perform an accurate phase comparison. FFT was used to calculate their spectra and extract the amplitude and phase. Finally, the phases of the main frequency components of the voltage and current were compared.
[0101] The formula for Fast Fourier Transform (FFT) and phase comparison can be expressed as follows:
[0102]
[0103] Where: X k is the frequency domain signal, x n is the time domain signal, N is the signal length, k is the index in the frequency domain, e is the base of the natural logarithm, and i is the imaginary unit.
[0104] For each frequency component X in the FFT result k , its phase φ k It can be calculated by the following formula:
[0105]
[0106] Among them, φ k =∠(X k ) where φ k represents the phase angle of the frequency component, ∠() represents the complex phase angle, Imag(X k ) and Real(X k ) are ∠(X k )’s imaginary and real parts.
[0107] When the input and output phase angles of the phase-locked loop are equal, phase locking is achieved. Calculate Δφ=φv[X k ]-φi[X k ], that is, the phase angle difference between the reference voltage and the actual current is the smallest. Using the mean square error (SME) indicator, the model parameters can be effectively adjusted to improve the prediction ability of the model:
[0108]
[0109] Objective function f 2 The difference between the reference current and the actual current is the smallest, and the mean square error (SME) indicator is also used:
[0110]
[0111] In the formula, i ref is the reference current i l is the actual current. Therefore, the final two objective functions are obtained:
[0112]
[0113] Step 5: Optimize the optimal parameters of the phase-locked loop and sliding mode controller by improving the NSGAII algorithm, and send the optimal parameters to the phase-locked loop and sliding mode controller. The phase-locked loop finally achieves phase locking after the integral link, so that the output current phase of the output inverter is equal to the grid voltage phase, and the sliding mode controller makes the output current equal to the reference current.
[0114] First, set the parameter range, where the phase-locked loop optimization parameter is c 1 and k, the optimized parameters of the phase-locked loop are K P and K I , and initialize the population: generate a population P containing N random solutions n , and for the population P n The obtained parameters are brought into the established model, and the fitness of the population is calculated according to the objective function established previously.
[0115] Using the SDR method to n Perform non-dominated sorting to form multiple fronts and calculate the crowding distance of each front. Then enter the main loop to check whether the termination criteria are met. If not, continue to perform the following steps:
[0116] According to the position and crowding distance of each solution in the non-dominated frontier, N parents are selected to form a new population P. n . And for P n Perform mutation operation to generate a new population P n . Calculate the new population P n The fitness of the population P is n Perform non-dominated sorting to form multiple fronts, calculate the crowding distance of each front, remove redundant solutions from the front with the worst crowding distance, and update the population.
[0117] Finally, a set of Pareto optimal solutions is obtained. A set of parameters closest to the coordinate origin is selected from the Pareto optimal solution set. This set of parameters can maintain the stability and performance of the system while meeting all control objectives.
[0118] Step 6: Figure 5 As shown, according to the output control rate, SPWM is used to control the full-bridge switch of the three-phase grid-connected inverter to achieve phase locking and current control.
[0119] First, for the output of the three-phase inverter, the current is transformed from the abc coordinate to the a-β coordinate through Clarke transformation, and the extended state observer model is established through the error between the output current and the reference current. Secondly, the parameters optimized by the improved NSGAII algorithm are introduced into the phase-locked loop and sliding mode controller. The phase-locked loop can adjust the frequency and phase of its output signal so that it gradually synchronizes with the grid voltage signal, and the phase difference remains constant; the sliding mode controller can obtain the total sliding mode control rate u by introducing the error e between the output current and the reference current and the sliding surface s into the sliding mode controller. conα and u con β Finally, u conα and u con β The two control rates are input into the SPWM modulation strategy to generate six PWM signals to control the switch control signals of the six switch devices, so that the DC output current tracks the given value.
[0120] Figure 6 The output current and grid voltage waveforms and tracking effects of phase A when the controller of the present invention is started under rated working conditions are shown in Table 1. Figure 6 It can be seen that by bringing the optimized parameters into the phase-locked loop and sliding mode controller, the current phase can track the grid voltage phase. When the system is stable, the output current waveform is stable and smooth, and the system has good stability.
[0121] Table 1 Experimental parameters
[0122]
[0123] The technical means disclosed in the scheme of the present invention are not limited to the technical means disclosed in the above-mentioned implementation mode, but also include technical schemes composed of any combination of the above-mentioned technical features. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also regarded as the protection scope of the present invention.
Claims
1. A multi-objective optimization method for sliding mode controller and phase-locked loop parameters based on grid-connected inverter, characterized in that: The following steps are involved: (1) Based on the filter inductor current and filter capacitor voltage of the three-phase grid-connected inverter, an equivalent system model of the three-phase grid-connected inverter in the α-β coordinate system is constructed; (2) Based on the equivalent system model, the sliding surface is designed and the sliding mode controller of the three-phase inverter is designed; (3) Design a phase-locked loop to ensure that the given phase of the grid-side current matches the phase of the grid voltage, so that the output current frequency and phase are synchronized with the grid voltage; (4) Propose an improved NSGAII algorithm to optimize the determined objective function; the improved NSGAII algorithm improves the diversity and convergence of the NSGAII algorithm by using a piecewise chaotic mapping initialization method and a linearly decreasing adaptive mutation operator and strengthening the dominance relationship; (5) The optimal parameters of the phase-locked loop and the sliding mode controller are optimized through the improved NSGAII algorithm, and the optimal parameters are sent to the phase-locked loop and the sliding mode controller. The phase-locked loop is phase-locked through the integral link, and the sliding mode controller outputs the control rate; (6) According to the output control rate obtained in step (5), the full-bridge switch of the three-phase grid-connected inverter is controlled by SPWM to achieve phase locking and current control.
2. The method for multi-objective optimization of sliding mode controller and phase-locked loop parameters based on grid-connected inverter according to claim 1, characterized in that: The equivalent system model on the α-β coordinate system in step (1) is: in, and They represent the current in the AC side inductance of the inverter on the α-axis and β-axis respectively. and They represent the voltage of the AC capacitor of the inverter on the α-axis and β-axis respectively, and u conα and u conβ K is the control voltage output by the inverter on the α-axis and β-axis; PWM is the inverter equivalent gain, which is used to convert the control input signal into the actual switching signal.
3. The method for multi-objective optimization of sliding mode controller and phase-locked loop parameters based on grid-connected inverter according to claim 1, characterized in that: The sliding mode controller of the three-phase inverter in step (2) is: Among them, u con is the control rate, i.e., the modulation signal of the inverter, which is used to control the switching action of the inverter and thus affect the output current; c1 is a positive coefficient of the sliding mode control, which is used to adjust the dynamic performance of the equivalent system; K is used to ensure the arrival condition of the sliding surface, i.e., the system state can reach and remain on the sliding surface; η is used to adjust the influence of the sign function to reduce chattering; sign(s) is the sign function, which is used to generate nonlinear control action to ensure that the system state slides on the sliding surface.
4. The method for multi-objective optimization of sliding mode controller and phase-locked loop parameters based on grid-connected inverter according to claim 1, characterized in that: The phase-locked loop in step (3) includes a phase detector, a loop filter, a voltage-controlled oscillator and a frequency divider; the phase detector compares the phase difference between the grid voltage and the internal synchronization signal of the control system, and converts the phase difference into a voltage signal; the voltage signal is processed by the loop filter and used to control the voltage-controlled oscillator, thereby adjusting the frequency and phase of the internal synchronization signal of the system to keep it synchronized with the grid voltage.
5. The method for multi-objective optimization of sliding mode controller and phase-locked loop parameters based on grid-connected inverter according to claim 1, characterized in that: The improved NSGAII algorithm in step (4) is: The initialization formula of Piecewise chaotic mapping is: Among them, P is a piecewise control factor used to divide the piecewise function, x i It is the state value generated by iteration, and its size is between 0 and 1. Due to its piecewise linear characteristics, PWLCM can theoretically provide better randomness and complexity, and can construct an initial population with a wide range, uniform distribution, and strong global search capabilities. The enhanced dominance relation SDR is used to solve the convergence and diversity of each iterative solution set; when a candidate solution x dominates another candidate solution y if and only if: in, Used to measure the degree of convergence of x, θ xy represents the acute angle between the target values of two candidate solutions, i.e., θ xy =arccos(f(x),f(y)); It is the size of the specific region of each candidate solution in the solution space; Adopt the linearly decreasing adaptive mutation probability operator strategy LDM to dynamically adjust the mutation probability: MP=MP max -(MP max -MP min )*(FE max -FE) / FE max In the formula, MP max is the maximum mutation probability, MP min is the minimum mutation probability, FE max is the maximum number of iterations, and FE is the current number of iterations.
6. The method for multi-objective optimization of sliding mode controller and phase-locked loop parameters based on grid-connected inverter according to claim 1, characterized in that: The objective function in step (4) is: The objective function f1 is to minimize the phase difference between the reference voltage and the actual current, that is, the phase-locked loop achieves phase locking when the input and output phase angles are equal. Calculate Δφ=φv[X k ]-φi[X k ], X k is a frequency domain signal; the objective function f2 is to minimize the difference between the reference current and the actual current; i ref is the reference current i l is the actual current.
7. The method for multi-objective optimization of sliding mode controller and phase-locked loop parameters based on grid-connected inverter according to claim 1, characterized in that: The implementation process of step (5) is as follows: (51) Set the parameter range, where the phase-locked loop optimization parameters are c1 and k, and the phase-locked loop optimization parameter is K P and K I ; and initialize the population to generate a population P containing N random solutions n , and for the population P n The solution in is normalized; the obtained parameters are brought into the established model, and the fitness of the population is calculated according to the objective function established previously; (52) Using the SDR method, the population P n Perform non-dominated sorting to form multiple fronts and calculate the crowding distance of each front; enter the main loop to check whether the termination criteria are met. If not, continue to execute (53); (53) According to the position and crowding distance of each solution in the non-dominated frontier, N parents are selected to form a new population P. n ; and for P n Perform mutation operation to generate a new population P n ; Calculate the new population P n The fitness of the population P is n Perform non-dominated sorting to form multiple fronts, calculate the crowding distance of each front, delete redundant solutions from the front with the worst crowding distance, and update the population; (54) A set of Pareto optimal solutions is obtained, and a set of parameters closest to the coordinate origin is selected from the Pareto optimal solution set to achieve that the output current phase of the output inverter is equal to the grid voltage phase, and the sliding mode controller makes the output current equal to the reference current.
8. The method for multi-objective optimization of sliding mode controller and phase-locked loop parameters based on grid-connected inverter according to claim 1, characterized in that: The implementation process of step (6) is as follows: Firstly, the output of the three-phase inverter is transformed from the abc coordinates to the α-γ coordinates through Clarke transformation, and the extended state observer model is established through the error between the output current and the reference current; secondly, the parameters optimized by the improved NSGAII algorithm are brought into the phase-locked loop and sliding mode controller, and the phase-locked loop adjusts the frequency and phase of its output signal so that it gradually synchronizes with the grid voltage signal and the phase difference remains constant; The sliding mode controller obtains the total sliding mode control rate u by bringing the error e between the output current and the reference current and the sliding surface s into the sliding mode controller. conα and u con β; Finally, u conα and u con The two control rates β are input into the SPWM modulation strategy to generate six PWM signals to control the switch control signals of the six switching devices, so that the DC output current tracks the given value.
Citation Information
Cited By
Inverter wireless parallel operation control system
CN121036189A
Inverter wireless parallel control system
CN121036189B
Terminal sliding mode phase synchronization control method and system based on improved particle swarm optimization
CN121770410A
A terminal sliding mode phase synchronization control method and system based on improved particle swarm optimization
CN121770410B
