Research on Similarity Matching of Grid Running Section Based on Stacked Autoencoder

Abstract

In order to utilize the large amount of historical data stored in the power system efficiently, and to provide the data support for the static security analysis, feature pattern extraction and situational awareness of the power grid, the idea of similarity matching of power grid running sections is proposed in this paper. The deep learning idea and model are introduced, and a similarity matching method for grid running section based on stacked autoencoder (SAE) is proposed. The algorithm process is divided into two stages which are layer-by-layer unsupervised pre-training and supervised fine-tuning. The effectiveness of the proposed method is validated on the IEEE 10-unit 39-bus system. The results show that the proposed method has high matching accuracy. In addition, the method greatly shorten the simulation time of training samples, with better performance and potential application value.