Optimal fast charging station locations for electric ridesharing service with online vehicle-charging station assignment

TL;DR

This paper introduces an integrated online vehicle-charging assignment model for electric ridesharing, optimizing fast charging station locations and operations to reduce delays and costs in a dynamic service setting.

Contribution

It presents a novel bi-level optimization framework combined with a surrogate-assisted approach for efficient planning and operation of charging infrastructure in electric ridesharing.

Findings

Charging delays reduced compared to existing methods

Up to 27.8% reduction in charging operation time with additional chargers

Effective online charging policy demonstrated on real-world scenario

Abstract

Electrified shared mobility services need to handle charging infrastructure planning and manage their daily charging operations to minimize total charging operation time and cost. However, existing studies tend to address these problems separately. A new online vehicle-charging assignment model is proposed and integrated into the fast charging location problem for dynamic ridesharing services using electric vehicles. The latter is formulated as a bi-level optimization problem to minimize the fleet's daily charging operation time. A surrogate-assisted optimization approach is proposed to solve the combinatorial optimization problem efficiently. The proposed model is tested on a realistic flexible bus service in Luxembourg. The results show that the proposed online charging policy can effectively reduce the charging delays of the fleet compared to the state-of-the-art methods. With 10 additional DC fast chargers installed, charging operation time can be reduced up to 27.8% when applying the online charging policy under the test scenarios.