Multistage Robust and Distributed Optimal Planning and Coordination of Power and Transportation Networks for Large-Scale EV Integration

Abstract

This thesis investigates the long-term planning and operational coordination challenges associated with large-scale electric vehicle (EV) integration in coupled power and transportation networks. As transportation electrification accelerates, EV charging demand exhibits significant spatial and temporal variability, creating strong interdependencies between traffic dynamics and power system operation. These characteristics make conventional planning approaches that treat power and transportation systems independently increasingly inadequate.

To address these challenges, this research develops a multistage planning framework that captures the sequential and irreversible nature of charging infrastructure investment under uncertainty. Robust and distributionally robust optimisation methods are incorporated to manage demand uncertainty without relying on precise probability distributions. The framework links long-term investment decisions with operational considerations across planning stages.

The thesis further examines the resilience of EV-integrated systems under disruptive events by embedding resilience-oriented objectives into the planning model. The results show that planning strategies based solely on expected conditions can lead to vulnerable charging infrastructure, whereas resilience-aware approaches improve system robustness and recovery capability.

In addition, distributed coordination mechanisms are proposed to support large-scale implementation under decentralised infrastructure ownership. Reinforcement learning is also explored as a complementary tool for short-term operational coordination. Case studies demonstrate that effective EV integration requires joint consideration of multistage planning, uncertainty management, power–transportation coupling, and scalable coordination mechanisms.