Optimizing autonomous electric taxi operations with integrated mobile charging services: An approximate dynamic programming approach

This paper focuses on optimizing the routing and charging schedules of an autonomous electric taxi (AET) system integrated with mobile charging services. In this system, a fleet of AETs provides on-demand ride services for customers, while mobile charging vehicles (MCVs) are deployed as a flexible complement to fixed charging stations, offering fast charging options for AETs. A dynamic programming model is developed to optimize the joint operations of AETs and MCVs, considering stochastics in customer demand, AET energy consumption, and charging station resources. The objective is to maximize the operator's overall profit over the entire planning horizon, including revenues from serving customer requests, travel costs, charging costs, and penalties associated with both fleets. To address the stochastic and dynamic nature of the problem, an approximate dynamic programming (ADP) approach, incorporating customized pruning strategies to reduce the state and decision space, is proposed. This approach balances immediate operational gains with future potential profits. A series of numerical experiments have been conducted to evaluate the effectiveness of the proposed model and algorithm. Results show that the ADP-based policy significantly improves system performance compared to classical myopic benchmarks.