Research on Operation Risk Prevention and Control Technology of Intelligent Distribution Network Based on Ultra Short Term Load Forecasting

Abstract

The distribution network is at the end of the power system and supplies power directly to the power users. With the development of intelligent distribution network, distribution network will be a complex intelligent system with multiple network structures, complex interconnection and mutual supply relationship, high penetration of distributed power sources, multi-objective, multi constraint and multi operation characteristics. The consequence of distribution system failure may lead to power failure from local to large area. According to statistics, about 80% of the power outage accidents are caused by the failure of the distribution system. Effective risk prevention and control based on the results of operation risk assessment of intelligent distribution network is an important prerequisite to ensure its safe and stable operation. The information obtained from the risk assessment of intelligent distribution network is very large. For online risk prevention and control, it is impossible to conduct online risk prevention and control analysis on all assessment results. It is necessary to select the most serious risk information from the massive result data, and then trigger online risk prevention and control measures to reduce the risk level to a certain extent. Based on ultra short term load forecasting, this paper proposes a risk prevention and control method for distribution network operation. Firstly, the state estimation of intelligent distribution network is based on the results of ultra short term load forecasting. Based on the results of state estimation, the risk assessment of intelligent distribution network is carried out. Secondly, based on the results of risk assessment, risk accidents are screened and sorted. Thirdly, the most important risk accidents are determined based on the network topology analysis to realize the online prevention and control of distribution network operation risk. Finally, the reliability and practicability of the algorithm are verified by the analysis of practical examples.