Energy Router Optimization Strategy Based on LSTM algorithm for Real-Time Congestion Prediction

As the core equipment of the energy Internet, energy routers (ERs) can optimize energy transmission and energy scheduling in time and space, thereby improving the transmission efficiency and stability of the multi-source power network. This paper establishes a graph model of the energy transmission network based on the topological structure of energy routing. Long and short-term memory (LSTM) is introduced to predict the congestion of ERs. As a result, an energy routing optimization strategy based on LSTM for real-time congestion prediction is proposed by considering the factors of congestion and transmission loss. Then Dijkstra algorithm is applied to determine the best path for energy transmission. Finally, numerical simulation verifies the feasibility and effectiveness of the energy routing optimization strategy proposed in this paper.