Research On Predictive Power Control Of PV Grid-connected Multivariable Model

Photovoltaic power generation is currently the fastest growing clean energy power generation technology,and the future development prospects are immeasurable.The grid-connected inverter is the core device of photovoltaic power generation and plays a decisive role in improving the stability of the system and suppressing the harmonics of the power grid.Therefore,the performance of the grid-connected inverter directly affects the grid-connected power quality.Inverter control technology is one of the important factors that determine the performance of inverters.As a new type of control strategy,Model Predictive Control(MPC)has gradually become a research hotspot.The model predictive control has good adaptability and robustness,and does not require inner loop current control and PWM modulation strategies,and the algorithm is easy to implement.This article first introduces the research status of photovoltaic power generation development and inverter control.Photovoltaic cell mathematical model and photovoltaic maximum power tracking technology were studied.Perturbation observation method was used to perform photovoltaic maximum power tracking control and simulation analysis was conducted.The mathematical model of the inverter is established under the three-phase stationary abc coordinate system and the two-phase stationary ?? coordinate system,and the instantaneous power of the photovoltaic inverter in the two-phase stationary coordinate system is analyzed,and the three-phase photovoltaic inverter is derived.Power forecasting model.In terms of control strategy,the feed forward decoupling control strategy for grid-connected inverters is first studied.Based on this,the model predictive power control is focused on.The dynamic response of feed-forward decoupling control depends on the PI controller and the dynamic response speed.Slower,and PWM modulation strategy generally uses SVPWM modulation,the algorithm is more complicated.Therefore,this paper studies the direct power control strategy based on the model ?? static coordinate prediction,and the simulation results show that the control strategy can quickly track the power changes,achieve decoupling control of output power,has good dynamic and static performance.In order to solve the problem of model predictive direct power control during the control process,such as delay in the control process and reduction of switching frequency,an improved multivariable model predictive power control strategy is proposed.A simplified two-step forecasting method is used to implement delay compensation,and is added to the cost function.The additional term S,which reduces the switching frequency,is designed to optimize the cost function and achieve multivariable control of the grid-connected inverter.This control can reduce the output power ripple,reduce the switching frequency,reasonably adjust the active and reactive output,and compensate the grid for reactive power.Finally,build a 5KW three-phase grid-connected experimental platform and complete experiments on the three-phase two-level inverter.A comparative analysis of the experimental results validates the effectiveness of the control strategies studied.