Research On The Key Problems Of Small Size Grid-connect Photovoltaic Power Generation System

Under the trend of the global energy structure changed to green renewable energy,the paper focuses on the distributed grid-connect photovoltaic power generationsystem, which is the most productive and the best prospects for development of theutilization of solar energy, and does some deep research on the key technical problems,mainly for the maximum power point tracking control, inverter output control andislanding detection. In order to ensure the grid-connect photovoltaic power generationsystem can be accessed to grid efficiently, stably, safely and with high quality, and toprovide the necessary technical support and engineering reference.The maximum power point tracking (MPPT) technique is the key to ensuremaximum power output of the grid-connect photovoltaic power generation system.After thorough analysis and comparison of several common control strategy of MPPT,aiming at the difficulty of the choice of step length, a new MPPT control strategybased on intermediate value of step length method combined with neural network isproposed. And the simulation results show that this strategy combines the dynamicand static characteristic advantage of the two methods, and has a more comprehensivesuperior control performance compared with the traditional INC method.Secondly, how to produce smaller error and total harmonic distortion (THD) ACsignal faster into the grid, that is the core issues need to be resolved by invertercontrol. After thorough analysis and comparison of several common control strategyof inverter output control, a hybrid control strategy is proposed based on the quasi-PRcontrol, repetitive control and PI control. And the simulation results show that thisstrategy has higher tracking accuracy and harmonic suppression capacity to meet therequirements of grid-connect.Finally, in order to avoid the generation of islanding, in the end of the paper, themodeling and simulation for islanding detection based on active frequency drift (AFD)method is conducted to verify the effectiveness ofAFD method in islanding detection.