Attack Detection based on Alternating Direction Multiplier Method for Distributed State Estimation in Smart Grid

Accurate state estimation is critical for smart grid monitoring. Due to the high degree of integration of cyber-physical systems, cyber attacks against various types of measurement data in smart grids are becoming more popular. As a result, the state estimation results for smart grid will be inaccurate, affecting the precise analysis and control of the smart grid’s current operating state. This paper proposes an ADMM-based distributed state estimate approach and a Markov attack detection mechanism to address the constraints of traditional state estimate and location of cyber-attacks. First, through separating power grid into many regions, we propose a distributed alternate direction state estimation approach based on multipliers that reduces the power system’s communication and computational expense. Second, by adopting presented combination attack model of false data injection attack and DoS attack, a Markov attack detecting mechanism based on distributed state estimate is developed. Finally, experimental results on the IEEE Bus-118 system demonstrate that our model has better efficiency and accuracy of suggested distributed state estimation and attack detection methods.