Electric Vehicles Reshape Urban Energy and Mobility Networks
The rapid rise of electric vehicles (EVs) is no longer just a shift in transportation technology—it is fundamentally transforming the way cities manage energy and mobility. As EV adoption accelerates globally, the integration of transportation and power systems has become a focal point for researchers, policymakers, and industry leaders. A recent comprehensive study published in IEEE Transactions on Smart Grid sheds new light on the complex interplay between electric vehicles, power grids, and urban transportation networks, offering a forward-looking framework for optimizing this evolving ecosystem.
Authored by Sheng Yujie, Guo Qinglai, Xue Yixun, Wang Jiawei, and Chang Xinyue from Tsinghua University and the Shanxi Institute of Energy Internet, the research presents a holistic cyber-physical-social systems (CPSS) perspective to model and optimize the coupling between power and transportation networks. The paper, titled “Collaborative Modeling and Optimization of Power-Transportation Coupling Network from Cyber-Physical-Social Perspective,” was published in 2024 in Automation of Electric Power Systems, with the digital object identifier (DOI) 10.7500/AEPS20230731006.
At the heart of the study is the recognition that EVs are not merely vehicles—they are mobile energy assets with the potential to serve as distributed storage units, capable of responding dynamically to grid conditions. Unlike traditional internal combustion engine vehicles, EVs can charge at various times and locations, offering unprecedented spatial and temporal flexibility. This flexibility, when properly harnessed, can significantly enhance grid stability, reduce peak load stress, and support the integration of renewable energy sources such as wind and solar.
However, unlocking this potential requires more than just advanced battery technology or fast-charging infrastructure. It demands a deep understanding of human behavior, real-time data exchange, and coordinated decision-making across multiple stakeholders—including utility companies, transportation authorities, charging operators, and individual drivers.
The Cyber-Physical-Social Framework
The authors argue that the conventional approach of treating EVs as passive loads or simple dispatchable resources is insufficient. Instead, they propose a three-layered CPSS framework that integrates physical network dynamics, social behavior modeling, and information-driven control strategies.
In the physical layer, the study models the transportation and power networks as interconnected systems where traffic flow and electric power flow influence each other through charging stations. Roads and charging facilities have finite capacities, and congestion—whether on highways or at charging hubs—can create bottlenecks that affect both mobility and energy distribution. The paper reviews various traffic network models, including static, semi-dynamic, and dynamic traffic assignment models, each suited for different time scales and planning horizons.
Static models, widely used in early studies, assume equilibrium conditions and are effective for long-term planning but fail to capture transient dynamics. Semi-dynamic models extend this by dividing time into discrete intervals, allowing for the simulation of time-varying demand and congestion propagation. Dynamic models, though computationally intensive, offer the most realistic representation of real-time traffic and charging behavior, making them ideal for operational decision-making.
The social layer addresses the critical yet often overlooked aspect of human decision-making. EV drivers do not behave as perfectly rational agents; their choices are influenced by personal preferences, risk perception, budget constraints, and psychological biases. The study emphasizes that traditional user equilibrium (UE) models, which assume homogeneous and fully rational behavior, are inadequate for capturing real-world complexity.
To address this, the authors advocate for the use of stochastic user equilibrium (SUE) models and behavioral theories such as cumulative prospect theory. These approaches account for bounded rationality, where individuals make suboptimal decisions based on incomplete information or cognitive limitations. For instance, a driver may choose a longer route to a less crowded charging station, even if it increases travel time, due to anxiety about battery depletion or queue length.
Empirical data plays a crucial role in refining these behavioral models. The researchers highlight the value of large-scale datasets from vehicle telematics, charging station logs, GPS trajectories, and even social media platforms. By fusing structured data (e.g., charging records) with unstructured data (e.g., user reviews), it becomes possible to build more accurate and interpretable models of driver behavior.
One of the key insights from the study is that user behavior is not static—it evolves over time in response to pricing signals, policy incentives, and technological advancements. For example, dynamic pricing at charging stations can shift demand away from peak hours, reducing strain on the grid. However, the effectiveness of such strategies depends on how well they align with user expectations and willingness to adapt.
The Role of Information and Multi-Agent Interaction
The information layer serves as the bridge between the physical and social components. It encompasses data collection, communication, and feedback mechanisms that enable real-time coordination. In modern urban environments, this layer includes smart meters, connected vehicle systems, navigation apps, and cloud-based energy management platforms.
The paper explores how different pricing mechanisms—such as time-of-use tariffs, congestion pricing, and incentive-based demand response—can be used to guide EV charging behavior. These mechanisms create a strategic interaction between multiple stakeholders: grid operators seek to minimize generation costs, transportation authorities aim to reduce traffic congestion, and charging service providers strive to maximize revenue.
This multi-agent environment often leads to complex game-theoretic dynamics. The authors analyze various pricing architectures, including Nash–Stackelberg–Nash games, where grid and traffic authorities act as leaders setting prices, while EV users respond as followers. In some cases, cooperative bargaining models are more appropriate, especially when the goal is to achieve a fair distribution of benefits among all parties.
A notable finding is that information asymmetry—where one party has more or better information than others—can significantly degrade system performance. For example, a charging operator may lack real-time data on traffic conditions, leading to suboptimal pricing decisions. Conversely, a navigation app may not have access to up-to-date grid load information, resulting in inefficient routing recommendations.
To overcome these barriers, the study calls for greater data sharing and interoperability across platforms. However, this raises concerns about privacy and cybersecurity. The authors suggest that privacy-preserving technologies such as federated learning, differential privacy, and homomorphic encryption could enable secure, decentralized computation without exposing sensitive user data.
From Theory to Practice: Challenges and Opportunities
While the theoretical foundations of power-transportation coupling are well established, practical implementation remains challenging. One major obstacle is the institutional fragmentation between energy and transportation sectors. In most cities, these systems are managed by separate agencies with different objectives, regulatory frameworks, and operational timelines.
The paper highlights the need for integrated planning and policy coordination. For example, urban planners should consider the placement of fast-charging stations not only in terms of traffic flow but also in relation to grid capacity and renewable energy availability. Similarly, utility companies must incorporate mobility patterns into their load forecasting models.
China has taken a proactive stance in this area. In December 2023, the National Development and Reform Commission released guidelines calling for enhanced integration between new energy vehicles and the power grid. The policy targets the establishment of five demonstration cities and over 50 bidirectional charging pilot projects by the end of 2025. These initiatives aim to test vehicle-to-grid (V2G) technologies, where EVs can feed electricity back into the grid during peak demand periods.
The authors note that while V2G holds great promise, its widespread adoption faces technical, economic, and behavioral hurdles. From a technical standpoint, frequent charging and discharging can accelerate battery degradation, raising concerns among vehicle owners. Economically, the revenue generated from grid services may not always outweigh the cost of battery wear. Behaviorally, drivers may be reluctant to allow third-party control over their vehicles’ charging schedules.
To address these issues, the study suggests a tiered approach. High-utilization fleets—such as taxis, ride-hailing vehicles, and delivery trucks—could be prioritized for V2G programs due to their predictable usage patterns and centralized fleet management. For private EV owners, less intrusive forms of demand response—such as time-based charging incentives—may be more acceptable.
Another emerging trend is the integration of wireless charging technology into road infrastructure. This could enable dynamic charging while vehicles are in motion, eliminating range anxiety and reducing the need for large battery packs. The paper references studies that explore the combined pricing of electricity and road usage in wireless charging scenarios, suggesting that such systems could lead to more efficient and equitable mobility solutions.
Resilience and Emergency Response
Beyond day-to-day operations, the coupling of power and transportation networks has important implications for disaster resilience. Natural disasters such as hurricanes, earthquakes, or wildfires can disrupt both energy supply and transportation routes. In such scenarios, EVs can play a dual role: as mobile power sources for emergency response and as flexible assets that can be rerouted to avoid damaged infrastructure.
The study reviews several models for coordinated emergency restoration, where power and transportation systems are jointly optimized to restore critical services. For example, mobile emergency generators can be pre-positioned based on predicted disaster impacts, while EVs can be directed to alternative charging locations if certain stations go offline.
The authors also warn of cascading risks—where a failure in one system triggers failures in the other. A blackout at a major charging hub could lead to traffic congestion as stranded EVs block roads. Conversely, a road closure could prevent repair crews from reaching a damaged substation. To mitigate these risks, the paper advocates for risk assessment frameworks that simulate interdependent failures and evaluate the effectiveness of contingency plans.
Future Directions
Looking ahead, the researchers identify several key areas for future investigation. One is the development of multi-scale, hierarchical optimization frameworks that can handle the vast computational complexity of integrated power-transportation systems. Directly coupling high-resolution models of both networks is often infeasible due to the sheer number of variables involved. Instead, the authors suggest using aggregation techniques—such as virtual power plants or mobility service providers—as intermediaries that simplify the interface between the grid and individual EVs.
Another promising direction is the use of artificial intelligence and machine learning to enhance predictive accuracy and decision-making speed. Reinforcement learning, in particular, shows potential for optimizing dynamic pricing and routing strategies in real time. However, the authors caution against over-reliance on black-box models that lack interpretability. Transparent, explainable AI systems will be essential for gaining public trust and regulatory approval.
Finally, the study emphasizes the importance of inclusive and equitable design. As EVs and smart charging become more prevalent, there is a risk of creating new forms of energy inequality. Low-income communities or rural areas may be left behind if infrastructure deployment is not carefully planned. Policies should ensure that the benefits of electrification are shared broadly across society.
Conclusion
The transformation of urban mobility and energy systems is well underway, driven by the convergence of electric vehicles, digital technologies, and sustainability goals. The work by Sheng Yujie, Guo Qinglai, Xue Yixun, Wang Jiawei, and Chang Xinyue provides a comprehensive roadmap for navigating this complex transition. By adopting a cyber-physical-social systems perspective, stakeholders can move beyond siloed thinking and embrace a more integrated, adaptive, and human-centered approach to urban infrastructure.
As cities around the world strive to reduce carbon emissions and enhance resilience, the lessons from this research offer valuable guidance. The future of transportation is not just electric—it is intelligent, interconnected, and deeply intertwined with the energy systems that power our lives.
Sheng Yujie, Guo Qinglai, Xue Yixun, Wang Jiawei, Chang Xinyue, Tsinghua University and Shanxi Institute of Energy Internet, Automation of Electric Power Systems, DOI: 10.7500/AEPS20230731006