Research On Control Strategy Of Photovoltaic Power Generation Micro-inverter Based On Neural Network

The large combustion of the fossil fuels not only make the non renewable energy source depletion, but also cause serious damage to the ecological environment. Solar energy, as a renewable clean energy, receives much attention. As one of the main ways of solar energy utilization, photovoltaic power generation has been widely used. In order to overcome the disadvantage of low output efficiency of centralized photovoltaic power generation system, each solar cell module is equipped with a miniature inverter to optimize the output efficiency of each module. Photovoltaic power generation system,which makes the whole output power maximum, is one of the trends in the future photovoltaic power generation. As one of the most important parts in the micro photovoltaic power generation system, the micro inverter has a direct influence on the efficiency and quality of the invert output. The advanced control algorithm is one of the main methods to improve the performance of the micro photovoltaic power generation system. In this paper, the neural network control method is studied and the method is used to the photovoltaic power generation micro inverter control system. The simulation results show that the control method has a good control effect.Firstly, the classification of neural network, neural network modeling method and neural network control structure are introduced. Then the overall design idea of neural network control system is summarized, which lays the foundation for the application of neural network in the micro inverter system for photovoltaic power generation.Secondly, based on the analysis of the solar photovoltaic power generation system,the model of the solar photovoltaic power generation micro inverter system and neural network control system are established according to the overall design, then the Matlab/Simulink is used to build the neural network control system simulation model which conclude grid-connected mode and island mode. The simulation experiment of the inverter system with neural network control algorithm is carried out under various operating conditions.The results show that the control algorithm can improve the quality of the output current and voltage. It also can reduce the harmonic content.Finally, the simulation model is built to verify the control strategy in the Matlab and PSIM software. The power level simulation circuit is built in PSIM, and the controller is built in Matlab. Simulation tests of various operating conditions are carried out on the simulation platform. A good control effect is obtained, and the feasibility of the control strategy is verified.