Design of the Relay Protection System Based on the Deep Learning Algorithm
In this paper,a new relay protection system integrating Convolutional Neural Networks(CNN)and Long Short-Term Memory(LSTM)is designed.The operation parameters of power system are obtained in real time by intelligent sensors,and the fault features are automatically extracted by deep learning algorithm.The performance of the system in fault detection,identification and location is verified by simulation experiments.The average fault detection time of the system is controlled within 20 ms,the accuracy of fault identification is 99.6%,and the error of fault location is controlled within 1%,which can quickly and accurately complete fault detection and diagnosis.The research results not only show the application potential of deep learning in the field of relay protection,but also provide new ideas for the development of smart grid.
万方数据
维普
