Situation Awareness Method for Electric Vehicles Charging Pile Clusters Based on Edge Computing

In response to challenges characterized by strong randomness, wide spatial distribution, and high power demand within short time frames by a large volume of Electric Vehicles (EVs), a situation awareness method for electric vehicle charging pile clusters is proposed based on edge computing. In this method, some factors such as EVs charging, EVs flow speed in road network are comprehensively considered by using edge computing, a physical framework is established for situation awareness electric vehicle charging pile clusters and some computational resources are reasonably allocated. In view of situation awareness for Electric Vehicles Charging Pile Clusters, an Isolation Forest(IF) Algorithm is used to perceive of the operational status of electric vehicles and the Variational Mode Decomposition(VMD) algorithm is used to understand the operational characteristics of electric vehicles. On this basis, the power demand of EVs charging is forecasted by Using Multilayer Perceptron(MLP) neural network model and the Gated Recurrent Unit(GRU) algorithm. Finally, the case is studied and the proposed method is feasible.