A command level anomaly detection model based on deep learning is proposed to solve the problems of poor evaluation effect,large amount of calculation and low execution efficiency in the network security evaluation of distributed charging piles.The sample data is trained based on the improved long short-time memory network,so as to improve the generalization ability and robustness of the model.The public subscriber/publisher model in the vehicle road collaborative charging pile network plat-form of State Grid Corporation is taken as an example to verify the proposed model.The simulation results show that the pro-posed model can achieve higher learning effect based on smaller learning samples,which can provide reference for the safe and reliable operation of distributed distribution network.
关键词
电力系统/充电桩/分布式配电网/网络异常/深度学习/马尔科夫时变模型
Key words
power system/charging pile/distributed distribution network/network anomaly/deep learning/Markovian time-varying model
维普
万方数据