Hierarchical coupled routing-charging model of electric vehicles, stations and grid operators

TL;DR

This paper presents a hierarchical model capturing the interdependent decision-making of EV drivers, charging operators, and grid managers, using a trilevel optimization framework to analyze impacts on urban transportation and electricity systems.

Contribution

It introduces a novel coupled routing-charging model with a trilevel structure, integrating transportation and electrical system decisions for EVs in urban environments.

Findings

Model effectively captures interdependencies among operators and EV behaviors.

Sensitivity analysis shows impact of EV penetration and incentives on system costs.

Algorithm successfully solves complex bilevel optimization problems.

Abstract

Electric Vehicles' (EVs) growing number has various consequences, from reducing greenhouse gas emissions and local pollution to altering traffic congestion and electricity consumption. More specifically, decisions of operators from both the transportation and the electrical systems are coupled due to EVs' decisions. Thus, decision-making requires a model of several interdependent operators and of EVs' both driving and charging behaviors. Such a model is suggested for the electrical system in the context of commuting, which has a typical trilevel structure. At the lower level of the model, a congestion game between different types of vehicles gives which driving paths and charging stations (or hubs) commuters choose, depending on travel duration and consumption costs. At the middle level, a Charging Service Operator sets the charging prices at the hubs to maximize the difference between EV charging revenues and electricity supplying costs, which are decided by the Electrical Network Operator at the upper level of the model, whose goal is to reduce grid costs. This trilevel optimization problem is solved using an optimistic iterative bilevel algorithm and simulated annealing. The sensitivity of this trilevel model to exogenous parameters such as the EV penetration and an incentive from a transportation operator is illustrated on realistic urban networks.