A Method for Eletric Vehicle Ownership Forecast Considering Different Economic Factors

Abstract

The construction of electric vehicles (EVs) charging station needs to be planed according to the ownership of EVs, traffic condition, population etc. Therefore a BP neural network based method to forecast the EV ownership for a city is presented in the paper, which considers the influence on the EV ownership caused by many related economy factors, including GDP of a city, vehicle production, per capita crude steel production, per capita generation capacity, road passenger traffic, highway mileage and the total population.  A BP neural network is set up for the forecast of EV ownership, and the input layer contains seven neutrons, which represent different economic factors. There are three neurons in its hidden layer, and the output is the EV ownership.  Then the method to predict the EV ownership of a city is presented, which is based on the forecast of the civilian car ownership in a city and the country. The EV ownership in the city of Chongqing from the year 2013 to 2020 is predicted, and the accuracy of the model is verified firstly, then the EV ownership in Chongqing is obtained, which is helpful to make plans for the development of electric vehicle.

DOI: http://dx.doi.org/10.11591/telkomnika.v11i4.2544