Abstract: In recent years,the intelligent power distribution systems have developed rapidly in the direction of data,networking and intelligence.With the increasing complexity of the cyberspace security situation,the network security of intelligent power distribution systems is facing enormous challenges.How to detect intrusion behavior against intelligent power distribution systems is a key technical issue that needs to be solved urgently.The intelligent power distribution system with control and protection switch as the terminal equipment was taken as the research object.The key algorithm of system intrusion detection was proposed.Three feature selection algorithms of Relief,MCFS and mRMR,were used to reduce the data respectively.The K nearest neighbor(KNN) algorithm was used as classifiers of the intrusion detection system,and the detection results were compared.The experiment results show that the intrusion detection algorithm proposed can effectively detect illegal attacks on intelligent power distribution systems.

Key words: intelligent power distribution system, intrusion detection, feature selection algorithm, K nearest neighbor(KNN) algorithm