Classification of Power Grid Operation States Based on DNN

Abstract

Aiming at the problems of low efficiency and subjectivity of current grid operation states classification methods of power grid operation states, a novel classification approach based on deep neural network (DNN) is proposed. Firstly, operation states of power grid are analyzed and summarized. Secondly, combining with the characteristics of the grid, the key indicators representing the operation states of the grid are pre-processed. Finally, DNN is introduced and used to deeply study the state of the historical grid to realize the classification of the current operation state. By extracting effective historical information, the method can realize the classification of the grid operation state, and assist the dispatcher to adjust the grid in time to avoid serious accidents. The simulation results of the New England 10-unit 39-bus system show that compared with the traditional BP neural network, the method proposed in this paper has higher accuracy. The model can help the dispatcher to quickly judge the state of the system so as to give a quick decision. The more accurate and effective macroscopic classification of the historical section can facilitate further data mining.