Robust Dynamic State Estimation of Power System Using Imperialist Competitive Algorithm

Abstract

Robust real-time state estimation in power systems is regarded as the first and foremost exigency in controlling and managing a safety network. Using prediction operations, dynamic state estimation (DSE) is considered as an applicable and efficient means for online tracking and monitoring of a network. In this paper, a new criterion has been introduced for DSE of power system to maximize the posterior probability density function. Furthermore, a comprehensive dynamic model consistent with information available from the network will be recommended. The dynamic state estimator proposed in this paper applies an imperialist competitive algorithm (ICA) to minimize the presented criterion, which is considered as constraints of the dynamic model of the network. The presented method is more advantageous than other state estimation methods, which are based on Kalman and statistical methods including independence to system linearity, Gaussian noise, and biopsy procedure. Moreover, according to considerations recommended for forming imperialists in ICA and predictable property of the suggested method, this state estimator is robust against data falsification. To compare, the presented method and other dynamic state estimators were implemented for an IEEE 9-bus system. Simulation results proved efficiency and superiority of the proposed method in robust estimation and fast tracking of network states.