ELM model for power system transient stability assessment
Zhang Li, Xiongwei Hu, Peng Li, Fang Shi, Zhihong Yu · 2017
The correct analysis of power system transient stability is of great significance to the safe and stable operation of power system and the construction of smart grid. Based on the basic theory of Extreme Learning Machine (ELM), this paper studies the transient stability of power system. Firstly, the simulation model is built to simulate power system for obtaining data sets. Then, the implementation of ELM is completed via programming on MATLAB, in which cross validation is used to optimize neural network structure. Recursive Feature Elimination is introduced in feature analysis and the evaluation method is improved to ensure the performance of the classifier. Finally, the stability prediction of an unknown system is given by ELM. The results show that ELM can make accurate transient stability assessment, and the training time is far less than SVM.