Research On The Charging Navigation Strategy Of Electric Taxis Under Complex Urban Roads

In the context of energy depletion and environmental pollution,it has become an irreversible trend for electric vehicles to replace fuel-powered vehicles.On the one hand,electric vehicles can greatly reduce the use of fossil fuels,reduce the dependence of energy supply on fossil fuels,and alleviate the dilemma of energy shortage.On the other hand,it can reduce the emission of carbon dioxide and other harmful substances and protect the environment.At present,a large number of cities have begun to popularize electric taxis and electric buses etc.However,there are still some problems in the process of promoting electric vehicles.First,There are endurance anxiety in almost electric taxi users,because they are worried that the battery of the electric taxi will run out of power during driving,which will affect their travel efficiency.Second,there are some problems in battery wear and tear,if electric taxi users often fully charge and discharge the battery,it is bound to reduce the service life of the battery and increase their travel expenses.Third,if a large number of electric taxis are charged at the charging station at the same time,the stability of the grid will be affected.At the same time,to reduce the electricity consumption of users near the load peak period,the grid will increase the electricity price during this period generally.If electric taxi users choose to charge near the load peak,it will inevitably increase the cost of electric taxi users.Given the three types of problems mentioned above,we urgently need to design a charging navigation strategy for charging path selection of electric taxis.The focus of this thesis is to propose a charging navigation strategy for electric taxis,which considers a variety of loss costs,including battery loss cost,charge and discharge loss cost,battery degradation cost,travel time cost,waiting time cost,and charge and discharge time.At the same time,we also takes into account the fact that electric taxis can be charged and discharged multiple times before reaching their destination.The purpose of this study is to minimize the comprehensive expenses of electric taxi users,improve the service life of the battery and guide users to discharge near the load peak and charge near the load valley,which play the role of shaving peaks and filling valleys.Thereby it can improve the stability for the grid.The main work of this thesis is as follows:To guide electric taxi users to discharge near load peaks and charge near load valleys,this thesis predicts the power load primarily and uses the predicted results to guide the pricing of electricity in route planning.To improve the accuracy of power load forecasting,this thesis proposes a forecasting model,named XGB-Kmeans-Bagging-LSTM.The simulation results show that the prediction model proposed in this thesis can fit the original power load curve effectively,and the prediction accuracy is higher than other algorithms.It guides the formulation of electricity prices in electric taxis route planning,so it can make route planning closer to the real scene.Immediately,aiming at the above problems encountered by electric taxis,a charging and navigation strategy for electric taxis that considers comprehensive loss and improves battery life is proposed.This thesis uses deep reinforcement learning DQN（Deep Q Network）to solve this problem.The simulation results show that the charging navigation strategy proposed in this thesis can find an optimal path for electric taxi users and minimize the comprehensive cost of electric taxi users.