Research On Control Strategies Of Single-Phase Grid-connected Photovoltaic Inverter Systems

This paper aims to research the the control strategy of single-phase photovoltaic grid-connected inverter system,so that the single-phase photovoltaic grid-connected inverter system has the ability of adaptive harmonic elimination to reduce the impact of grid-connected current distortion on the power system.In order to improve the single-phase photovoltaic grid-connected inverter system,this paper has done the following work:1)Firstly,the basic structure and principle of the two-stage single-phase photovoltaic grid-connected system is introduced.The front-stage boost circuit and the last-stage single-phase bridge inverter circuit consist the main circuit.Establish the mathematical model of photovoltaic cell and front-stage boost circuit,analyze its power-voltage output characteristic curve,study the commonly used maximum power point tracking methods such as disturbance observation method,constant voltage tracking method and conductance increment method.The appropriate MPPT control method is selected to enable the front-stage boost circuit to output a stable DC voltage for the inverter.2)Aiming at the serious phenomenon of grid-connected current distortion,the selective characteristic harmonic elimination strategy is applied to the grid-connected inverter based on the two-stage single-phase photovoltaic grid-connected inverter system.Optimize the grid-connected current waveform and improve current tracking capability.The double loop PI control strategy and the QPR control strategy are mainly studied in depth.Double loop voltage and current PI control strategy.Firstly,the virtual quadrature component is constructed,and the double loop proportional integral(PI)control of the single-phase dq synchronous coordinate system is realized by Park transformation.Then a harmonic feedforward compensation link is added.Based on the FFT principle and the method of amplitude phase synthesis,a harmonic detection system is designed to detect low-order characteristic harmonics with high content,and then the PWM modulation wave is reconstructed by harmonic injection to achieve grid connection,improve current tracking optimization.QPR control strategy.The QPR control can form resonance at the fundamental frequency,and realizes no-difference tracking of the tracking amount,which can avoid the error of the PI controller tracking.Therefore,the QPR control is selected as the controller of the inverter system,which constitutes a new transfer function for the high odd-order harmonic frequency,and then uses the harmonic detection system to extract the corresponding characteristic harmonics of the error current and compensate it up.3)Apply BP neural network to the harmonic compensation system.For the double loop voltage and current PI control strategy,in the compensation system,the LM-BP neural network is used to adaptively adjust the compensation gain of the characteristic harmonics.The total distortion rate(THD)is reduced from 3.52% of the traditional double loop voltage-current PI control strategy to 1.32%,but as for the condition of sudden load changes,the inverter will be unstable due to excessive adjustment time.For the QPR control strategy,the method with adaptive learning rate is used to improve the BP neural network adjusting the compensation gain,reducing the THD from 3.42% of the traditional QPR control strategy to 2.06%,and when the load suddenly changes,the THD is reduced from 5.32% of traditional QPR control strategy to 3.10%.