Agent-Based Modeling and Simulation for Systematic Operations of Shared Automated Electric Vehicles
Fugen Yao,Xiqun (Michael) Chen,Panagiotis Angeloudis,Wenwen Zhang
DOI: https://doi.org/10.1061/9780784482933.211
2020-01-01
Abstract:This paper proposes a framework of future-oriented agent-based modeling and simulation (ABMS) for various operational scenarios and optimization of shared automated electric vehicles (SAEVs). We establish an efficient scheduling algorithm between vehicles and passengers, and real-time matching algorithm for vehicles and charging stations. The scheduling algorithm includes two processes. First, each customer finds a candidate vehicle, and then the platform performs the final scheduling. The ABMS framework simulates the complicated matching relationship and interactions among the on-demand ride services platform, passengers, vehicles, and charging stations. Field operations of a large fleet of SAEVs are implemented using the real ride-sourcing order data in the road network of Hangzhou, China. The simulation results under different scenarios are comprehensively compared. The sensitivity of several critical parameters is analyzed, e.g., the SAVE fleet size, recharge mileage, and charging speed. The proposed ABMS modeling framework can be extended to incorporate a variety of vehicle types, and support decision making of advanced vehicle scheduling strategies, pricing, and relocation.