Research On High-precision Load Forecasting And Demand Response Strategies For Residential Users

With the continuous improvement of the living standard of urban and rural residents in China,the proportion of the residential load in the social load structure is gradually increasing.At the same time,as the peak load of the power grid increases year by year,how to ensure the long-term safe and stable operation of the power system at the peak load period has become an important problem to be solved urgently.Residential user load has the characteristics of large quantity and wide distribution.On the basis of residential user load forecasting,effective integration and utilization of residential user demand response resources can play a key role in alleviating the power grid operation pressure.On the basis of investigating the current situation of load forecasting and demand response of residential users,this paper studies the high-precision load forecasting method and demand response strategy.The main work completed is as follows:(1)Based on the detailed residential user load database,the load characteristics of residential users are comprehensively analyzed from different scales(single and aggregate households)and time scales(short-term and long-term),and the impact of external factors such as weather and date types on residents are studied.The demand response ability of various types of household electricity load are excavated through data curve analysis and other means,then this paper selects temperature-controlled load as the main target of the residential user demand response strategy.(2)On the basis of dimensional reduction processing of the research data,the combined deep learning network short-term load forecasting model is built combining with the traditional artificial neural networks(BP,ELM)and deep learning network(LSTM),and the accuracy of the forecasting model is verified through the analysis of the case study.The results show that the proposed forecasting model meets the high precision forecasting standard,and also shows excellent performance in the proposed new index evaluation system for demand response scheduling.(3)Based on the first-order equivalent ETP model of air conditioning load,the actual parameter model of air conditioning load is established by integrating historical load and environmental temperature data,and the effects of air conditioning load participating in demand response are analyzed from two aspects of different wheel control time interval and participation degree.The demand response module is introduced into the load forecasting model,and the demand response strategy of air conditioning load based on load forecasting is proposed and analyzed by the case study.The results show that the proposed method realizes the participation of residents in demand response and achieves the purpose of reducing peak load on the basis of ensuring high-precision day-ahead load forecasting.