[Paper Review] Joint Optimization of Autonomous Electric Vehicle Fleet Operations and Charging Station Siting

Justin Luke, Mauro Salazar+2 more|arXiv (Cornell University)|Jun 30, 2021

Electric Vehicles and Infrastructure13 references27 citations

TL;DR

This paper proposes a convex joint optimization framework that simultaneously determines the optimal siting and charging infrastructure for autonomous electric vehicle (AEV) fleets and their macroscopic operations. Using a linear program integrating network flow models with charging station design, the approach minimizes total fleet costs—including routing, charging, procurement, and demand charges—achieving up to 10% reductions in total cost, peak charging load, and empty vehicle travel, with a shift toward high-power Level 2 AC stations and reduced DC fast charging capacity.

ABSTRACT

Charging infrastructure is the coupling link between power and transportation networks, thus determining charging station siting is necessary for planning of power and transportation systems. While previous works have either optimized for charging station siting given historic travel behavior, or optimized fleet routing and charging given an assumed placement of the stations, this paper introduces a linear program that optimizes for station siting and macroscopic fleet operations in a joint fashion. Given an electricity retail rate and a set of travel demand requests, the optimization minimizes total cost for an autonomous EV fleet comprising of travel costs, station procurement costs, fleet procurement costs, and electricity costs, including demand charges. Specifically, the optimization returns the number of charging plugs for each charging rate (e.g., Level 2, DC fast charging) at each candidate location, as well as the optimal routing and charging of the fleet. From a case-study of an electric vehicle fleet operating in San Francisco, our results show that, albeit with range limitations, small EVs with low procurement costs and high energy efficiencies are the most cost-effective in terms of total ownership costs. Furthermore, the optimal siting of charging stations is more spatially distributed than the current siting of stations, consisting mainly of high-power Level 2 AC stations (16.8 kW) with a small share of DC fast charging stations and no standard 7.7kW Level 2 stations. Optimal siting reduces the total costs, empty vehicle travel, and peak charging load by up to 10%.

Motivation & Objective

• To address the gap between isolated optimization of AEV fleet operations and charging station siting by jointly modeling both problems.
• To minimize total cost for an E-AMoD operator, including fleet procurement, routing, charging, and infrastructure costs.
• To quantify the impact of EV type (e.g., range, size, cost) on optimal infrastructure and operational design.
• To demonstrate that joint optimization yields significantly better outcomes than optimizing operations on a pre-existing, baseline station layout.

Proposed method

• Formulates a network flow model that represents vehicle movements, charging events, and battery state-of-charge across space, time, and SoC levels.
• Integrates charging station siting decisions into the optimization by modeling discrete charging rates (e.g., Level 2, DC fast) at candidate locations.
• Constructs a linear program that jointly optimizes fleet size, routing, charging schedules, and infrastructure procurement costs.
• Incorporates electricity demand charges and allows for charge throttling and self-loop arcs to represent origin-destination trips within the same region.
• Uses a time-expanded graph with discrete time steps and battery charge level discretizations to model dynamic vehicle operations.
• Solves the resulting large-scale convex optimization problem using off-the-shelf solvers (e.g., Gurobi), achieving global optimality.

Experimental results

Research questions

• RQ1 How does jointly optimizing AEV fleet operations and charging station siting affect total system cost compared to sequential optimization?
• RQ2 What is the optimal mix of charging station types (e.g., Level 2 vs. DC fast) and their spatial distribution for an E-AMoD fleet?
• RQ3 How do different EV models (with varying range, size, and cost) influence the optimal infrastructure and operational design?
• RQ4 To what extent does joint optimization reduce peak charging load and empty vehicle travel compared to a baseline based on current-day station siting?
• RQ5 How does the autonomy of the fleet enable more distributed and efficient charging compared to human-driven EVs?

Key findings

• The joint optimization reduces total system cost by 9.59% compared to a baseline scenario based on current-day charging station siting.
• The optimal charging station configuration features a 64% reduction in DC fast charging capacity, favoring high-power Level 2 AC stations (16.8 kW) over standard 7.7 kW Level 2 stations.
• Peak charging load is reduced by 10.07%, and empty vehicle travel (rebalancing distance) is reduced by 11.20% due to more spatially distributed and accessible charging infrastructure.
• Smaller, lighter, and cheaper EVs (e.g., Leaf S) with high energy efficiency are the most cost-effective in terms of total ownership cost, despite limited range.
• The optimal siting is more spatially distributed than current-day siting, with the top 13 stations in the optimized scenario covering 52% of zones to achieve 74% of total capacity, compared to just 4 zones in the baseline.
• The optimization problem with 5.3 million variables was solved in 2.33 hours using a 24 vCPU, 64GB RAM instance, demonstrating computational feasibility.