Application And Research Of Neural Network Inverse Control In Single-phase Photovoltaic Inverter System

For its own advantages of solar power, it has become one of the main energy for human to develop and apply. Photovoltaic power generation is the main form of solar energy utilization and the inverter as the most important component in photovoltaic power generation system, its performance directly affects the efficiency and quality of the inverter. Exploring the advanced control algorithm is the primary means to improve and enhance the performance of a single-phase inverter system. This paper mainly studied the neural network inverse control method and applied it to the single-phase inverter control system designed in this paper.Firstly, this paper describes the concept and principle of the artificial neural network and inverse system and related theory of inverse control method. It respectively introduces the topology of single deviation-grid photovoltaic inverter and flyback micro grid-connected photovoltaic inverter and analyses its mathematical model. Based on that, the reversibility of above two inverter system is studied.Secondly, the neural network inverse model of the single deviation-grid photovoltaic inverter and flyback micro grid-connected photovoltaic inverter is established. Combined with classic PI control algorithm, two simulation models based on neural network inverse compound control algorithm about inverter systems are separately set up by Matlab/Simulink. Simulation experiments focus on using the neural network inverse compound control algorithm of single-phase off-grid inverter for a variety of working conditions, as well as comparing its inverter control effect with ones under circumstance of pure PI control algorithm.Finally, the single-phase off-grid photovoltaic experimental platform was designed and the design of the hardware circuit and software were introduced in detail. Meanwhile the method to implement neural network special arithmetic unit based on FPGA was introduced. Combining with inverter, it constitutes a single-phase off-grid photovoltaic inverter system based on neural network inverse composite control algorithm. Inverter experiments were done and the experimental results are recorded finally based on above. Simulation and experiment showed that this method is feasible and had a better control effect.