Traffic-power flow coordination and EV dynamic wireless charging infrastructure planning toward sustainable and resilient electrified transportation network

Abstract

Electric vehicles (EVs) are gradually replacing traditional internal combustion engine vehicles as a more sustainable and environmentally friendly mode of transportation. Traditional EV charging methods rely on wired connections, which introduce significant challenges (e.g., inconvenience of plugging and unplugging chargers, long charging wait times, and range anxiety). This hinders the broader adoption of EVs to some extent. Emerging dynamic wireless charging (DWC) technology can effectively address these issues by enabling wireless electricity transmission from power systems to in-motion EVs using the magnetic resonance coupling principle. The success of numerous DWC pilot projects around the globe demonstrates the promising prospects for the widespread implementation of DWC infrastructure in the near future.

As renewable energy penetration grows and extreme weather events occur more often, ensuring sustainability and resilience has emerged as a central goal for managing power systems under normal and extreme conditions, respectively. The introduction of DWC load as a new type of load poses additional challenges to resilient and sustainable power system operations. In addition to operational considerations, it is also necessary to reasonably plan the DWC infrastructure for convenient charging and more optimal electrified transportation network performance. The decision-dependent uncertainty (DDU), i.e., travelers’ stochastic path selection behavior influenced by deployed DWC infrastructure, brings huge challenges to planning. Motivated by these factors, this thesis will investigate three crucial aspects: sustainable power system operation, resilient power system operation, and DWC infrastructure planning under multiple sources of uncertainties.

Firstly, to understand the characteristics of DWC loads, an equivalent circuit analysis of a typical DWC system with multiple segmented transmitting coils is presented. Based on this analysis, a more accurate trapezoidal power model for a single EV is developed. Utilizing this model, an aggregated EV DWC load model is proposed, considering road traffic flow characteristics and headway characteristics. Moreover, to cope with uncertain risks, a sustainable operation strategy leveraging the conditional value at risk (CVaR) framework is put forward.

Secondly, a semi-dynamic traffic assignment approach with heterogeneous vehicles is developed to optimize vehicle flow allocation for enhancing grid resilience. Damage from extreme weather events is accounted for by representing power line failures as outages and traffic road disruptions as decreased capacity. The proposed service restoration model considers the coordination of multiple sub-recovery tasks, including EV semi-dynamic routing, bidirectional DWC, power distribution network topology reconfiguration, and power distribution network power dispatch, to enhance power system resilience.

Finally, a robust co-planning framework is developed to boost flexibility and cost efficiency of electrified transport networks. This model incorporates dynamic wireless charging lane (DWCL) deployment, bus selection, and remote-controlled switches (RCS) installation. Travelers' stochastic path choice behavior, modeled using the random utility principle, captures the decision-dependent relationship between travelers' path choice probabilities and deployed DWCLs. A robust optimization framework with two stages and three levels is developed to tackle decision-independent uncertainties (DIU), including stochastic traffic demand and renewable energy source outputs. An inexact nested column-and-constraint generation (INCCG) algorithm with McCormick relaxation and heuristic sequential bound tightening is proposed to accelerate the solution process.