Modeling Research Of Flexible Amorphous Silicon Photovoltaic Power Generation System

Solar energy plays an important role in all kinds of renewable energy with its clean, no pollution, huge reserves as well as other advantages. The flexible amorphous silicon thin film photovoltaic generation technology has been a typical representative of new energy technology since its inception in 1976 with its advantages of less materials, short recovery cycle, more annual power generation, and so on. The modeling research of this paper is carried out just based on this type of photovoltaic cell and its power generation system, whose purpose is to lay theoretical foundation for application technology development of this kind of solar cell. The following work has been carried out in detail:1. A circuit model of flexible amorphous silicon photovoltaic cell is proposed. This study adopted quasi Newton iterative method and self-organizing migration method to extract all unknown parameters of the circuit model, thus establishing a single diode circuit model, according to the I-V data of flexible amorphous silicon solar cell tested at standard solar light. By comparing simulation experiment data with test data, the result shows that standard deviation of the circuit model is 0.79%, and the residual error is 0.05%. The study lays the core foundation for further optimization of flexible amorphous silicon photovoltaic cell power generation performance.2. Besides, a model of lithium battery which is for energy storage in the flexible amorphous silicon photovoltaic power generation circuit has been built. The circuit model is simple and accurate, which is built based on experiment data. Then, charging process simulation and validation experiment are also carried out.3. The Boost circuit and MPPT control module are selected to build the simulation model of flexible amorphous silicon photovoltaic power generation system in Simulink, MATLAB.4. For key problems of low output efficiency of flexible amorphous silicon photovoltaic power generation system when environmental factors change, this study proposes a maximum power tracking control algorithm combined the PID module with BP neural network, which solved the problems such as a long tracking time (0.04s), large overshoot (11.7%), and large steady-state error (0.5V) caused by the PID control and variable step disturbance observation method. Simulation result shows that, the improved control algorithm can finish the flexible amorphous silicon solar cell maximum power point tracking within 0.01s, the overshoot is decreased to 0.003%, and the steady state error is reduced to 0.15V. In conclusion, the improved algorithm can eliminate the voltage oscillation during transient operation and improve the robustness of the system.