On the Fair-Efficient Charging Scheduling of Electric Vehicles in Parking Structures

This work examines the off-line electric vehicle (EV) scheduling problem for cloud based parking operators, that apriori accept parking reservations for EVs requesting charging services during their stay. Specifically, it examines the fair EV charging scheduling problem, where fairness refers to the achievable charging levels of EVs contending for energy utilities within a planning horizon. For finding fair utility allocations the α-fairness approach is used, inspired by welfare economics, that is formulated as an integer linear program (ILP) and as an ant colony optimization (ACO), considering both the system’s and EV owners’ constraints and requirements. It is shown that with this approach the operator is able to control the fairness-efficiency trade-off (with system efficiency affecting the operator’s revenue) by appropriately selecting the inequality aversion parameter  to best meet targeted performance metrics. Further, it is shown that ACO, deriving near-optimal allocations, significantly outperforms the ILP-based algorithm in terms of processing time (up to 99%), thus it is a promising approach when optimal ILP allocations cannot be derived fast enough for a practical implementation.