New Strategy Optimizes Power-Traffic Systems by Accounting for Traffic Delays
As electric vehicles (EVs) continue to surge in popularity, the integration between urban power grids and transportation networks has become increasingly intricate. The rise of electric mobility is no longer just about replacing internal combustion engines with battery-powered drivetrains—it’s about redefining how energy and transportation systems interact. A groundbreaking study published in the Proceedings of the CSU-EPSA introduces a novel approach to managing this complex relationship by addressing a critical yet often overlooked factor: traffic delay.
The research, led by Liu Jue, Mu Yunfei, Dong Xiaohong, and Jia Hongjie from Tianjin University and Hebei University of Technology, presents a comprehensive strategy for optimizing the coordinated operation of power and traffic systems—specifically, how EV charging behavior is influenced by real-world traffic conditions. The team’s work focuses on the concept of “traffic delay factor” and its impact on the spatial and temporal distribution of EV charging demand. By integrating this dynamic into a unified optimization model, the researchers demonstrate how both grid stability and traffic flow can be improved simultaneously.
The Growing Complexity of Power-Traffic Coupling
The convergence of power distribution networks and urban transportation systems—commonly referred to as Power-Traffic Coupling Systems (PTCS)—has emerged as a key area of study in smart city development. As millions of EVs connect to the grid for charging, their energy demands create fluctuating loads that can strain distribution infrastructure, especially during peak hours. At the same time, the movement of these vehicles across city roads affects traffic congestion, travel times, and overall network efficiency.
Traditionally, EV charging strategies have been based on static assumptions: drivers choose routes and charging stations based on current traffic and pricing information at the moment of decision. However, in reality, traffic conditions are not static. There is a time lag—known as traffic delay—between when a driver makes a routing decision and when they actually experience the road conditions. This delay can lead to suboptimal choices, such as arriving at a charging station only to find it overcrowded, or encountering unexpected congestion that increases travel time and energy consumption.
Liu Jue and her team recognized that ignoring this temporal discrepancy undermines the effectiveness of existing coordination strategies. “Most current models assume perfect information,” explains Mu Yunfei, one of the co-authors. “But in real-world scenarios, traffic data is always slightly outdated. If we don’t account for that delay, our optimization models are built on flawed premises.”
A Forward-Looking Model Based on Traffic Delay Factor
To address this gap, the researchers developed a two-part framework. The first component is a spatiotemporal prediction model for EV charging demand that incorporates the concept of the traffic delay factor. This factor quantifies the time lag between traffic data collection and actual vehicle movement, allowing the model to simulate more realistic travel patterns.
The model begins by simulating individual EV behavior using travel chain analysis and Monte Carlo sampling. Each vehicle is assigned a battery capacity, initial state of charge, departure time, and destination based on statistical distributions derived from real-world data. As the simulation progresses, the system evaluates when an EV’s remaining battery level or driving range falls below a threshold, triggering a charging need.
At this point, the traffic delay factor comes into play. Instead of assuming that the current traffic conditions will persist, the model predicts how traffic will evolve over the next several minutes. It calculates the expected travel time for each possible route by factoring in not just current congestion levels, but also how those levels are likely to change due to delayed traffic waves. This is achieved by analyzing historical traffic flow data and identifying optimal delay values that maximize the correlation between upstream and downstream traffic patterns.
By doing so, the model can recommend a route that minimizes not only distance but also the risk of encountering worsening congestion. This refined path selection process leads to more accurate predictions of when and where EVs will arrive at charging stations, which is crucial for the second part of the framework: the coordinated operation optimization model.
Minimizing Total System Cost Through Integrated Decision-Making
The second component of the proposed strategy is an optimization model designed to minimize the total operational cost of the PTCS. This cost includes four main elements: power grid losses, traffic network congestion, charging station operational expenses, and EV user costs such as travel time, queuing, and electricity charges.
Rather than optimizing each component in isolation, the model takes a holistic approach. When an EV identifies a charging need, the system evaluates all feasible charging stations within its driving range. For each candidate station, it simulates the entire charging process—from leaving the current location, navigating through the traffic network, waiting in line, to completing the charge and reconnecting to the grid.
The simulation accounts for dynamic variables such as real-time traffic speeds, queue lengths at charging stations, and fluctuating electricity prices. The pricing mechanism itself is adaptive: when a charging station approaches its load capacity, it implements a congestion-based pricing scheme that increases the cost slightly to discourage additional arrivals. This price signal is fed back into the decision-making process, encouraging EVs to choose less crowded alternatives.
The objective is to select the charging station that results in the lowest total system cost across all stakeholders. This means balancing the need for efficient grid operation (reducing power losses and voltage fluctuations) with smooth traffic flow (minimizing congestion and travel time) and reasonable user experience (avoiding long waits and excessive charges).
“This isn’t just about saving money,” notes Dong Xiaohong. “It’s about creating a more resilient and sustainable urban infrastructure. When EVs charge at the right place and time, everyone benefits—the grid, the roads, the charging operators, and the drivers.”
Simulation Results Show Clear Advantages
To validate their approach, the research team conducted a detailed case study using a 24-node urban traffic network coupled with a modified IEEE 33-bus distribution system. The simulation involved 5,000 EVs making trips throughout a 24-hour period, with charging stations located at key intersections and connected to the medium-voltage grid.
Three scenarios were compared:
- Scenario 1: No traffic delay considered, fixed pricing, and independent operation of power and traffic systems.
- Scenario 2: No traffic delay considered, but dynamic pricing and coordinated operation.
- Scenario 3: Traffic delay factor included, dynamic pricing, and full coordination.
The results were compelling. In Scenario 3, the total system cost was reduced by 12.7% compared to Scenario 1 and by 6.3% compared to Scenario 2. The reduction came from multiple sources: lower power losses in the distribution network, decreased congestion on major roadways, shorter queuing times at charging stations, and reduced travel costs for EV users.
Notably, the inclusion of the traffic delay factor led to a more balanced distribution of charging demand across stations. Without it, many EVs were routed to the same few stations based on outdated traffic data, leading to localized congestion both on the roads and in the grid. With the delay-aware model, charging loads were spread more evenly, reducing peak stresses on both systems.
The impact on traffic performance was particularly striking. Using the Traffic Performance Index (TPI) as a metric, the researchers found that high-congestion segments (TPI > 0.8) decreased by 28% in Scenario 3 compared to Scenario 1. Even when compared to the coordinated but delay-agnostic Scenario 2, the improvement was significant, especially during morning and evening rush hours.
Similarly, power losses in the distribution network were reduced by 15.4% in the delay-aware scenario. This was largely due to better load balancing—EVs were encouraged to charge during off-peak hours and at locations where the grid had sufficient capacity, avoiding voltage drops and thermal overloads.
Implications for Smart Cities and Grid Management
The implications of this research extend far beyond academic interest. As cities worldwide push for electrified transportation to meet climate goals, they face the dual challenge of managing increased electricity demand and maintaining efficient urban mobility. Traditional grid planning and traffic management systems are ill-equipped to handle the bidirectional interactions between energy and transportation.
The strategy proposed by Liu Jue and her colleagues offers a practical solution. By embedding traffic delay into the decision-making loop, it enables more accurate forecasting and proactive control. This is especially valuable for distribution system operators (DSOs) who must ensure grid reliability while accommodating unpredictable EV loads.
Moreover, the model supports the development of intelligent charging networks that can respond dynamically to real-time conditions. Charging station operators can use similar frameworks to optimize pricing and capacity planning, while city planners can integrate these insights into broader transportation policies.
From a policy perspective, the research underscores the importance of data integration. Accurate traffic delay estimation requires access to high-resolution traffic flow data, which may involve collaboration between transportation agencies, navigation service providers, and utility companies. The success of such a system depends on open data standards and interoperable platforms.
Future Directions and Practical Deployment
While the current study focuses on daytime public charging, the authors acknowledge that future work should expand to include home charging, especially overnight. Residential charging patterns are influenced by different factors—such as household schedules and time-of-use tariffs—and integrating them into a unified model presents new challenges.
Additionally, the team plans to explore mixed charging networks that include both fast and slow chargers, as well as vehicle-to-grid (V2G) capabilities. These extensions would allow for even greater flexibility in load management and could further enhance system efficiency.
Practical deployment of this strategy would require integration with existing smart grid and intelligent transportation systems (ITS). Real-time data from traffic sensors, GPS devices, and smart meters would feed into the optimization engine, which could be hosted in a cloud-based control center. The output—a set of recommended charging stations and routes—could then be delivered to drivers via mobile apps or in-car navigation systems.
User acceptance will be key. While the model aims to minimize total system cost, individual drivers may prioritize convenience or cost savings over system-wide efficiency. Therefore, incentive mechanisms—such as discounted charging rates or priority access—may be necessary to encourage compliance.
A Step Toward Truly Integrated Urban Infrastructure
The work of Liu Jue, Mu Yunfei, Dong Xiaohong, and Jia Hongjie represents a significant advancement in the field of energy-transport integration. By recognizing that traffic delay is not just a nuisance but a fundamental characteristic of urban mobility, they have developed a more realistic and effective approach to managing the complex interplay between EVs, roads, and power grids.
Their model demonstrates that small improvements in prediction accuracy—such as accounting for a few minutes of traffic delay—can lead to substantial gains in system performance. This aligns with the broader trend in smart city development: leveraging fine-grained data and advanced analytics to create more responsive, efficient, and sustainable urban environments.
As EV adoption continues to accelerate, the need for such integrated solutions will only grow. Cities that invest in coordinated power-traffic management systems today will be better positioned to handle the challenges of tomorrow—reducing emissions, improving air quality, and enhancing the quality of life for their residents.
In conclusion, this research is not just about optimizing EV charging. It’s about reimagining how urban systems operate as interconnected wholes rather than isolated parts. By bridging the gap between transportation and energy, the team has laid the groundwork for a smarter, more resilient future.
Published by Liu Jue, Mu Yunfei, Dong Xiaohong, and Jia Hongjie from Tianjin University and Hebei University of Technology in the Proceedings of the CSU-EPSA, DOI: 10.19635/j.cnki.csu-epsa.001411.