Voltage Safety Evaluation And Prediction Of Microgrid Using Chaotic Time Series And RBF Neural Network

With the global energy shortage and the worsening environment,the research and development of microgrids has gradually been unanimously recognized at home and abroad.Microgrid refers to a kind of independent and autonomous micro power system that can integrate energy storage devices,distributed energy sources,loads,energy monitoring devices,protection devices into a self-managing,protection and control.The load and distributed power in the microgrid can be used flexibly and efficiently to the greatest extent,and can solve the problem of grid connection of some distributed power.Uncertainty of intermittent energy access in the microgrid,flexible loads such as electric vehicles,and random faults in system lines and equipment threaten the safe operation of the microgrid.In response to this problem,this paper takes safety as the base point,discusses the voltage safety of the microgrid and estimates the voltage safety of the microgrid in the short time scale and the long time scale.The research results have certain reference significance for the planning design and protection control of the microgrid.This article mainly includes the following work content:1.A microgrid voltage safety assessment model is proposed: identifying and processing the historical abnormal data of the voltage at the public connection point in the microgrid to ensure the validity and integrity of the data,and then combined with the volatility and intermittent characteristics of the wind turbines and photovoltaic output in the microgrid,based on the voltage safety evaluation index algorithm,it is determined whether the microgrid is in a safe state.2.Proposed a short time scale prediction model of micro-grid voltage safety: judging the chaos of the pre-processed micro-grid voltage data to determine whether the micro-grid voltage data is in a chaotic state.On the basis of chaos,the state space reconstruction of the data,and the short-term voltage safety prediction model of the weighted first-order local microgrid is established by fusing the neighboring points of the interval.3.Proposed a long-term prediction model of micro-grid voltage safety: combining the Auto Regression model with the RBF neural network to form an improved RBF-AR neural network model.Based on the weighted first-order local method of the chaotic time series merging the adjacent points of the interval and the improved RBF-AR neural network prediction model,the combined effective degree weight coefficient of the nonlinear microgrid voltage time series is solved to estimate the microgrid voltage Safety.The prediction results are compared with the prediction results of the single weighted first-order local method and the RBF-AR neural network model,and it is concluded that the combined model has better stability for the security prediction and reduces the risk ability of the single model.