Smart Charging Strategy Balances Grid Stability and EV Owner Savings
As electric vehicle (EV) adoption accelerates globally, urban power systems face mounting pressure from uncontrolled charging behaviors. The surge in EV ownership, projected to reach 60 million in China by 2030 according to industry forecasts, is transforming residential energy consumption patterns. However, this transition brings significant challenges, including increased peak load demands, amplified load fluctuations, and potential overloading of distribution transformers. These issues not only threaten the stability and safety of local power grids but also lead to higher electricity costs for consumers. In response, researchers have been exploring strategies to manage EV charging more intelligently, aiming to align vehicle charging with grid capacity and renewable energy availability. A recent study published in Power System Protection and Control presents a novel approach that successfully addresses both grid-level and user-level concerns, offering a practical solution for the evolving energy landscape of smart communities.
The core challenge lies in the inherent conflict between the needs of the power grid and those of individual EV owners. From the grid operator’s perspective, the primary objective is to maintain stability by minimizing load fluctuations, particularly the difference between peak and off-peak demand, known as the peak-to-valley difference. Uncoordinated EV charging, especially when it occurs during already high-demand periods, exacerbates this problem, creating a “peak-on-peak” scenario that can push transformers beyond their capacity limits, as highlighted in the study’s simulation of an uncontrolled charging scenario. This overloading reduces equipment lifespan and increases the risk of power outages. Conversely, EV owners are primarily motivated by economic factors, seeking to minimize their charging costs. The most effective way to achieve this is by charging during off-peak hours when electricity prices are lowest, typically during the night. While this user-centric approach reduces individual expenses, it can inadvertently create a new peak in demand during the low-tariff period, undermining the grid’s stability goals. Previous research has often focused on optimizing for one of these objectives—either grid stability or user cost—leading to solutions that are effective in isolation but suboptimal for the system as a whole. Some studies have incorporated renewable energy sources like solar power, but many fail to integrate energy storage or consider the computational challenges of managing large fleets of EVs in real time. The research by Kang Tong, Zhu Jiran, Feng Churui, Fan Min, Ren Lei, and Tang Haiguo from the State Grid Hunan Electric Power Company Limited Research Institute and Chongqing University directly confronts these limitations with a comprehensive and innovative framework.
The study introduces a dual-layer multi-objective optimization model designed for a photovoltaic-storage-charging integrated community. This model is a significant departure from single-objective strategies, as it simultaneously pursues two critical goals. The first layer, referred to as the “grid layer,” is tasked with minimizing the community’s load peak-to-valley difference. This objective ensures that the total power demand on the local transformer remains within safe operational limits, preventing overloads and promoting a stable, reliable power supply for all residents. The second layer, the “user layer,” focuses on minimizing the charging costs for individual EV owners. This dual focus is crucial for creating a strategy that is not only technically sound but also economically viable and attractive to end-users. A key innovation is the hierarchical relationship between these two layers. The solution from the grid layer—specifically, the optimized charging load profile—acts as a hard constraint for the user layer. This means that while users are given the freedom to optimize their charging schedules for the lowest cost, they must do so within the boundaries established by the grid’s stability requirements. This elegant design ensures that the pursuit of user savings does not compromise the safety and integrity of the power distribution network. The model is further enhanced by its integration of a 200-kilowatt photovoltaic (PV) system and a 200-kilowatt-hour energy storage system. This allows the community to maximize the on-site consumption of solar energy, reducing reliance on the main grid and further lowering overall energy costs.
To solve this complex, multi-constrained optimization problem, the researchers employed a cutting-edge metaheuristic algorithm known as the Rat Swarm Optimizer (RSO). Traditional optimization methods often struggle with the high dimensionality and non-linear constraints of such models, frequently getting trapped in local optima that are not the best possible solution. The RSO algorithm, inspired by the hunting behavior of rats, is designed to maintain a dynamic balance between local exploitation (intensive search around a promising solution) and global exploration (searching new, unexplored areas of the solution space). This capability makes it particularly well-suited for finding a globally optimal or near-optimal charging schedule that satisfies all the model’s constraints. The algorithm works by simulating a population of “rats” that iteratively update their positions in a search space, representing different possible charging schedules. The “rats” are guided by the position of the current best solution, allowing the entire population to converge toward an optimal strategy. The use of this advanced algorithm is a testament to the complexity of the problem and the sophistication of the proposed solution, enabling the researchers to find a charging plan that effectively balances the competing demands of grid operators and consumers.
A critical component of the study’s practicality is its proposed cloud-edge collaborative scheduling architecture. This architectural design directly addresses a major bottleneck in large-scale EV management: computational scalability. Traditional centralized control systems, where all data is sent to a central cloud server for processing, face severe challenges as the number of connected EVs grows. The sheer volume of data and the complexity of the optimization calculations can overwhelm the central server, leading to long response times and slow decision-making, which is unacceptable for real-time energy management. The researchers’ solution decentralizes the computational load. In their architecture, the computationally intensive grid-layer optimization is performed on the cloud side, which has vast processing power. The results of this grid-level optimization—essentially the target load profile—are then sent to edge-side devices, such as smart fusion terminals installed within the community. These edge devices, which are closer to the actual EVs and chargers, then perform the user-layer optimization locally. This means that the final, personalized charging schedule for each EV is calculated on the community’s local network, drastically reducing the amount of data that needs to be transmitted back and forth and minimizing latency. This cloud-edge collaboration leverages the strengths of both systems: the cloud’s massive computing power for system-wide optimization and the edge’s proximity and speed for local, real-time decision-making. This architecture is scalable, efficient, and resilient, making it a viable model for future smart grid deployments.
The researchers conducted a detailed case study using a simulated community in Hunan, China, to validate their proposed strategy. The simulation involved 150 EVs, a 200-kilowatt PV system, a 200-kilowatt-hour battery storage system, and a 1000-kilovolt-ampere (kVA) distribution transformer. They compared the performance of five distinct charging scenarios: uncontrolled charging in a standard community, uncontrolled charging in a photovoltaic-storage-charging community, grid-layer-only ordered charging, user-layer-only ordered charging, and their proposed dual-layer multi-objective strategy. The results were compelling and demonstrated the clear superiority of their integrated approach. Under the uncontrolled charging scenario, the community’s peak load reached 970.89 kilowatts, exceeding the transformer’s 900-kilowatt capacity and putting it in an overloading state. Even with the presence of solar and storage, uncontrolled charging in the integrated community still resulted in a peak load of 925.66 kilowatts, which was still above the safe limit. This starkly illustrates that simply adding renewable energy and storage is not enough without intelligent control.
The dual-layer strategy achieved remarkable improvements. Compared to the uncontrolled charging in the integrated community, the proposed method reduced the load peak-to-valley difference by an impressive 40.47%. This dramatic reduction means a much more stable and manageable load profile for the grid operator, effectively eliminating the risk of transformer overloading. The peak load was brought down to 827.25 kilowatts, well within the safe operating range. This is a critical achievement for ensuring the long-term reliability and safety of the distribution network. The benefits for consumers were equally significant. The strategy reduced the average charging cost by 52.63%. This was accomplished by strategically shifting charging to periods of low electricity prices and high solar generation, allowing residents to take full advantage of the community’s on-site renewable energy. When compared to single-layer strategies, the dual-layer approach proved to be the most balanced and effective. While a user-layer-only strategy achieved the lowest individual charging cost, it risked creating a new peak in demand during the night, which could destabilize the grid. Conversely, a grid-layer-only strategy successfully flattened the load curve but did so without regard for user costs, potentially reducing consumer willingness to participate. The dual-layer model successfully navigated this trade-off, achieving performance that was close to the best in both categories without the drawbacks of either extreme.
The implications of this research extend far beyond a single community in Hunan. It provides a robust, scalable blueprint for the future of urban energy management in the age of electrified transportation. As cities worldwide strive to meet climate goals and increase the penetration of renewable energy, the integration of EVs into the power grid will be a defining challenge. This study demonstrates that with the right combination of advanced optimization algorithms, a multi-objective design philosophy, and a modern cloud-edge computing architecture, it is possible to turn the challenge of EV charging into an opportunity. The opportunity is to create a more resilient, efficient, and economical energy ecosystem. The dual-layer model ensures that the grid remains stable and secure, which is a non-negotiable requirement for any utility. At the same time, by demonstrably reducing consumer costs, it fosters user engagement and acceptance, which is essential for the widespread adoption of any new technology. The success of this strategy hinges on its holistic view, recognizing that a truly smart grid must serve both the infrastructure and the people who rely on it. The research by Kang Tong et al. represents a significant step forward in making the vision of a sustainable, integrated energy future a practical reality.
The study also acknowledges its limitations, noting that it did not account for variations in EV models, battery types, or charging behaviors beyond the assumed patterns. Future work will aim to incorporate these factors to enhance the model’s realism and applicability. Nevertheless, the core framework of a dual-layer, cloud-edge optimized strategy provides a powerful foundation. It sets a new standard for EV charging research by moving beyond single-objective optimization and addressing the critical issue of computational scalability. As the number of EVs on the road continues to grow exponentially, solutions like this will be indispensable for building the smart, sustainable cities of tomorrow. The findings offer valuable insights for utility companies, urban planners, and technology developers, providing a clear path toward a future where electric vehicles are not a burden on the grid, but a key component of a smarter, more flexible, and more sustainable energy system.
Kang Tong, Zhu Jiran, Feng Churui, Fan Min, Ren Lei, Tang Haiguo, State Grid Hunan Electric Power Company Limited Research Institute, Chongqing University, Power System Protection and Control, DOI: 10.19783/j.cnki.pspc.230998