Multiperiod Planning for Renewable Electric Vehicle Charging Stations in Coordinated Power Distribution and Transportation Networks With Non-Anticipativity Constraints

This article proposes a planning model for renewable electric vehicle charging station (REVCS) to meet the growing demand for electric vehicles (EVs) in the context of the net-zero transition. Focusing on the multiperiod strategic deployment and device capacity optimization of REVCS in the coordinated urban transportation network (UTN) and power distribution networks (PDNs), the model is formulated by a multistage robust optimization (RO) problem, which makes decisions based on the uncertainties of the present period for adhering to non-anticipativity constraints. To efficiently solve the planning model, an improved robust dual dynamic programming (RDDP) method is proposed, which decomposes the multistage RO problem into upper and lower approximation problems using approximate convex hulls and Lagrangian hyperplanes, respectively, and iteratively refines solutions through a forward and a backward pass process until optimal. Numerical experiments on a 33-bus PDN coupled with 12-node UTN and a larger 54-bus PDN coupled with 25-node UTN systems validate the multiperiod planning model and proposed solution methodology.