The Urban Recharging Infrastructure Design Problem with Electric Taxi Demand Prediction Using Convolutional Neural Network

Abstract

In this study we first design a convolutional long short-term memory model that predicts taxi drop-off and pick-up demands for time periods in urban areas. Then, based on the predicted taxi demand in drop-off and pick-up hotspots, a mathematical model is proposed to optimize the location of recharging stations so as to minimize the locating cost of stations and delay cost for recharging service.