Research On Key Technologies Of Multi-source Microgrid Based On Model Predictive Control

In recent years,human beings have faced dual pressures caused by energy shortages and environmental degradation.With the deepening of pressures,human beings are seeking new breakthroughs,that is,the research on multi-source microgrid systems based on new energy sources Attention.As a new type of power supply system,the multi-source microgrid system has the advantages of diverse power supply forms,high utilization rate of renewable energy,low environmental pollution,flexible operation mode,and high reliability of power supply.This paper analyzes and studies the system management problems and model predictive control methods of a multi-source microgrid system that uses wind energy,hydrogen production from electrolyzed water,hydrogen-oxygen fuel cells,and coal-to-methanol as basic microsources.The main work is as follows:First,based on the analysis of the operating characteristics of each micro-source,the corresponding mathematical model is studied.Secondly,to study the variability of wind power output,that is,in order to better study the impact of variable wind power output on system management,the improved BP neural network model,improved GA-BP and improved PSO-BP combined forecasting model are used.The comparative analysis of the forecast effect of wind power shows that the improved PSO-BP combined forecast model has the best forecast performance.Then,comprehensively considering the economics and power supply reliability of the new energy multi-source microgrid system,the MPC control method is proposed for this system management problem.The local wind energy local maximum absorption and the main grid energy interaction minimum are used as the optimization objective function,based on the MPC strategy,Build SSM prediction model,use genetic optimization algorithm for rolling solution,aim at the interference and prediction error generated by the system during operation,introduce feedback mechanism to embed it in the MPC framework,and then use rolling time domain method to the system The interference is compensated,and the realization process of MPC strategy is given.Finally,based on the actual measurement data of the four seasons in the Northeast and the Netherlands,the first and second cases were analyzed.By comparing the wind power absorption effect of the system during operation,it was verified that the proposed method can effectively reduce the interactive power.Achieve maximum consumption of local wind power.