Generation Method of Power Network Security Defense Strategy Based on Markov Decision Process

Abstract

Aiming at the problem that the current generation method of power networksecurity defense strategy ignores the dependency relationship between nodes,resulting in closed-loop attack graph, which makes the defense strategy notgenerate attack path, resulting in poor defense effect and long generationresponse time of power network security defense strategy, a generationmethod of power network security defense strategy based on Markov decisionprocess is proposed. Based on the generation of network attack and defensediagram, the paper describes the state change of attack network by usingMarkov decision-making process correlation principle, introduces discountfactor, calculates the income value of attack and defense game process,constructs the evolutionary game model of attack and defense, solves theobjective function according to the dynamic programming theory, obtains theoptimal strategy set and outputs the final results, and generates the powernetwork security defense strategy. The experimental results show that the proposed method has good defense effect and can effectively shorten thegeneration response time of power network security defense strategy.