Power Router Optimization Strategy Based On Real-time Congestion Prediction

With the increasing proportion of sustainable energy connected to the grid,the scale and structure of the traditional power network are not enough to complete a large number of new energy integration work.Due to the intermittently,randomness and volatility of renewable energy,traditional power grids cannot effectively connect these increased renewable energy sources to the grid.As a large amount of renewable energy generation is connected to the public grid,it will inevitably lead to fluctuations in power,frequency and line voltage.Energy Internet technology based on electricity is an important way to realize energy transformation.Energy Internet can realize extensive access to new energy,so as to solve the uneven distribution of energy,as well as the imbalance of time dimension and space dimension.As the core switching device of the energy Internet,the energy router connects the whole power grid.It has the functions of energy distribution,regulation,scheduling,management,etc.,and provides guarantee for the security,stability and reliability of the energy network.Therefore,in the work of energy Internet,line loss and node congestion in the process of energy transmission should be considered comprehensively to improve energy transmission efficiency and reduce energy loss.It has important theoretical significance to promote the development of energy Internet.Aiming at the transmission congestion problem in the energy Internet,the congestion degree of the routing node is defined considering the relationship between the current power throughput and the maximum power throughput of the power router.Based on the defined congestion degree,the Long and Short Term Memory(LSTM,Long and Short Term Memory)and LSTM prediction model based on IGWO(IGWO,Improved Grey Wolf optimization)to predict the congestion of power router.By applying entropy weight method and CRITIC weight method respectively,the predicted congestion degree and line loss indicators were weighted,and the power router optimization strategy based on LSTM prediction model and the power router optimization strategy based on IGWO-LSTM prediction model were proposed.The main research contents are as follows:(1)Based on the fact that the real-time power throughput is greater than the maximum power throughput of the node,the congestion of the routing node will result in the scheduling difficulties of the energy Internet and even the network paralysis.In view of the congestion problem of the power routing node in the energy Internet,this paper considers the constraint relationship between the maximum throughput and throughput,and gives the congestion degree formula to describe the congestion degree of the routing node.To pave the way for the power router optimization strategy.(2)Based on the defined congestion degree,LSTM is introduced to predict the congestion of power router.The two indexes affect the optimal path selection,and the entropy weight method is introduced to weight the two indexes.Then the optimal transmission path is determined by Dijkstra algorithm using weighted index.Then a real-time congestion prediction power router optimization strategy based on LSTM is proposed.(3)Considering that the prediction effect of LSTM is affected by learning rate,LSTM number of layers and regularization parameters,an improved grey Wolf optimization algorithm is used to optimize the habit rate,LSTM number of layers and regularization parameters,and a long and short memory model based on improved grey Wolf optimization is proposed to predict the congestion of electric power routing nodes.Considering that the entropy weight method does not take into account the relationship between two indicators and affects the relevant weight,the CRITIC weight method is introduced to weight congestion and transmission loss,and the Dijkstra algorithm is adopted to obtain the best path.Then,an optimization strategy of power router based on IGWO-LSTM prediction model is proposed.(4)Aiming at the proposed power router optimization strategy based on LSTM prediction model,the topology structure of 14 routing nodes is taken as an example to verify the feasibility and effectiveness of the power router optimization strategy.Aiming at the proposed power router optimization strategy based on IGWO-LSTM prediction model,taking the topology structure of 10 routing nodes as an example,the advantages and effectiveness of the proposed power router optimization strategy are verified by simulation.