Research On Location Planning Of Electric Taxi Charge/Exchange Station Based On Data Mining

With the adjustment of the world’s energy industry structure and people’s attention to environmental issues,electric vehicles have been widely promoted due to their clean and efficient functions.The construction of charging infrastructure is an important part of large-scale promotion of electric vehicles.Planning electric vehicle charging infrastructure is conducive to the healthy development of electric vehicles.As the main category of electric vehicles,electric taxis have more frequent charging behaviors than ordinary electric vehicles.Reasonable planning of their charging infrastructure is an important guarantee for the development of the electric taxi industry,and the prediction of the time and space distribution of the charging load is the electric taxi charging.Prerequisites for facility planning and site selection.Based on a large amount of data,this paper studies the space-time distribution and prediction method of electric taxi charging load,and plans and selects the location of electric taxi charging infrastructure.The main research contents are as follows:First of all,A grid method for analyzing the charging load of electric taxis is proposed.The grid analysis of the time and space in the studied area is used to obtain the space-time distribution of the charging load in the studied area.In-depth study of the spatial distribution and spatial distribution of electric taxis.The time distribution of the charging load on weekdays and the charging load on holidays is studied separately.The research results show that the time distribution of the charging load of electric taxis during holidays is quite different from that of the working day.Secondly,A fully connected neural network is used to predict the charging load.A three-layer fully connected neural network is used.The input layer includes time,gridded longitude,and gridded latitude.The output layer is the charging demand in the grid.The error between the output value and the actual value of the network training can quickly converge.The accuracy of the method is verified by comparing the predicted value of the spatial distribution of the charging load with the actual value at a certain moment.Finally,A planning and location model for centralized charging/exchanging power stations and decentralized charging piles for electric taxis is proposed.The model takes into account the charging demand distribution of electric taxis,the characteristics of the planning area,the construction cost of charging/exchanging power stations and charging piles,and electric rental.The distance between the vehicle and the charging pile and the charging/replacement station,and the load capacity of the distribution network are subject to the objective function of minimizing the construction cost of the charging/reforming station and the charging pile and the minimum distance between the electric taxi and the charging/replacement station and the charging pile.Using Haikou’s online car data sources,data mining was performed on taxi order latitude and longitude,driving distance,and charging requirements.The DBSCAN algorithm was used to solve the planning of the distributed charging piles,and the planning and location of the distributed charging piles were obtained.The results show that the number of sites for decentralized charging piles is relatively stable.However,some areas still need to be supplemented by centralized charging/exchanging stations.The K-means algorithm was used to plan and select the centralized charging/exchanging station.The results show that when 22 centralized charging/exchanging stations are constructed,the total objective function value is the smallest.And divide the centralized charging/exchanging station service and deployment area.