Distributed Coordination of Electric Vehicles in Unbalanced Distribution Grids: Enhancing Resilience to Peer-to-Peer Communication Failures

In this paper, we propose coordinated charging and discharging of Electric Vehicles (EVs), enabled by a distributed optimization-based approach that is scalable and resilient to communication failures. The underlying optimization problem is formulated to maximize the economic benefits for grid-connected EVs, subject to transformer thermal constraints and nodal voltage constraints enforced on an unbalanced distribution grid. Our distributed coordination approach is underpinned by a consensus-based alternating direction method of multipliers (ADMM) that encompasses an iterative negotiation process amongst neighboring EVs. The charge-discharge profiles are computed by EVs locally (and asynchronously) using limited information exchange via peer-to-peer communications. We prove that our proposed EV coordination algorithm converges, even in the event of communication failures such as packet losses and random disconnections of EVs. Numerical simulations with the IEEE 13-node test feeder, incorporating empirical real-life data of EVs and residential loads, show that the proposed approach yields a 78% reduction in operational costs compared to uncoordinated EV charging, while also preserving network operational limits. In the presence of communication failures, the proposed approach resulted in an 8 times slower convergence speed compared to an ideal communication network.