Research On Control Strategy Of Inverter Based On Improved Swarm Intelligence Algorithm

As a key device for the connection of the distributed generation system and the power grid,the grid-connected inverter plays an important role in improving the overall performance of the system and reducing the switching harmonics.Because the LCL filter has strong filtering performance and has the ability to suppress harmonics,it is widely used in grid-connected inverter systems.Since the high-order part of the LCL inverter will cause the system to be unstable,the system needs to be effectively controlled in the actual control process.Aiming at those problem above,the nonlinear control strategy for LCL structural inverter circuits is designed in this paper.This strategy not only has the stability as that of traditional control scheme,but also obtains good control accuracy.In the research field of control strategy,the selection of controller parameters is a technical problem.In order to solve this problem,this paper proposes a Knitting Optimization Algorithm(KOA)to optimize the parameters of the controller.Due to the characteristics of systemization,rapidity,diversity and strong global optimization ability,the KOA provides a reliable and practical parameter selection method for the controller.The work of this paper is arranged as follows:Firstly,the single-phase LCL inverter topology is constructed with the theoretical basis and technical indicators.In the dq rotating coordinate system,the mathematical model of nonlinear control strategy based on LCL inverter system is established.The nonlinear control strategy consists of a Differential Tracker(DT),a State Observer(SO),and a Nonlinear Controller(NC).This strategy improves the control accuracy and response speed of the system through the integrated control of the various orders of the control target.Secondly,through the improvement of update method and optimization mode,a KOA is proposed.The KOA is based on the idea of swarm intelligence algorithm,combining the advantages of other optimization algorithm.KOA provides a practical and efficient method for the selection of controller parameters.The algorithm with system simulation model is established to verify the effect of the KOA combined with controller.The performance of KOA is analyzed and compared with Genetic Algorithm(GA)、 Simulated Annealing Algorithm(SAA)and Particle Swarm Optimization(PSO).The controller parameters are the objects to be optimized,and the optimization process uses the appropriate criteria to calculate the fitness function.According to the simulation results,the applicability and superiority of the KOA are verified comprehensively and systematically.Finally,an experimental platform for DC/AC full-bridge single-phase LCL inverter circuit system is constructed.The nonlinear control strategy based on KOA is compared with PR and PID control strategy by simulation and experiment.The feasibility of the nonlinear control strategy and the effectiveness of the KOA are verified by simulation and experimental results.