Smart DC Microgrid Cuts EV Charging Costs by 28%
The rapid rise of electric vehicles (EVs) is reshaping the global transportation landscape, driven by environmental concerns, technological advancements, and supportive government policies. As EV adoption accelerates, one of the most pressing challenges for consumers remains the cost of charging. While the environmental benefits of EVs are clear, the financial burden of electricity, especially during peak hours, can deter potential buyers and strain household budgets. In response to this growing concern, a team of researchers from the State Grid Zhejiang Marketing Service Center and the Beijing Key Laboratory of Demand Side Multi-energy Carriers Optimization and Interaction Technique has developed a novel control strategy that significantly reduces EV charging costs through a smart, coordinated DC microgrid system.
Published in the November 2024 issue of Electric Drive, the study introduces an innovative approach that integrates photovoltaic (PV) solar power, energy storage systems (ESS), and the conventional power grid into a unified DC microgrid. This system is specifically designed to optimize energy flow, minimize reliance on expensive grid electricity, and maximize the use of self-generated renewable energy. The results are compelling: the proposed strategy can reduce daily charging costs by up to 28.1% compared to existing methods, and by a staggering 62.4% compared to standard public charging stations. This breakthrough has the potential to make EV ownership more affordable and accelerate the transition to a low-carbon future.
The core of this research lies in its sophisticated coordination of multiple energy sources. Traditional EV charging often relies solely on the grid, exposing users to fluctuating electricity prices, particularly during peak demand periods. While solar panels and home batteries offer a solution, their integration is often simplistic, failing to account for the dynamic interplay between energy generation, storage, and consumption. The team, led by Wang Chaoliang, recognized that a more intelligent, real-time control system was needed to navigate the complexities of variable solar output, unpredictable EV charging patterns, and time-of-use electricity tariffs.
The proposed system is built on a common DC bus architecture, a design choice that enhances efficiency by eliminating the need for multiple AC/DC conversions. The key components—PV arrays, a lithium-ion battery bank, the AC power grid, and the EV charging load—are all connected to this central DC bus through specialized power converters. The true innovation, however, is not in the hardware but in the software and control logic that orchestrates the entire network.
A fundamental aspect of the strategy is the dynamic prediction of EV charging demand. The researchers employed the Monte Carlo algorithm, a powerful statistical method, to model the highly random nature of EV usage. This includes variables such as daily driving distance, which follows a log-normal distribution based on real-world travel surveys, and the time of day when a vehicle is plugged in, which also exhibits a probabilistic pattern. By simulating thousands of possible scenarios, the algorithm generates a highly accurate forecast of the aggregate charging load over a 24-hour period. This predictive capability allows the system to anticipate energy needs and make proactive decisions, rather than simply reacting to immediate demand.
To manage the system’s operation, the researchers developed a comprehensive framework based on six distinct operational modes and four voltage bands. This dual-layered approach ensures both energy efficiency and system stability. The six modes represent different energy flow scenarios, dictated by the real-time balance between solar generation and charging demand, as well as the state of charge (SOC) of the battery.
For instance, in “Mode I,” solar production is so high that it not only meets the EV charging load but also fully charges the battery to its upper SOC limit. At this point, the excess solar energy is automatically fed back into the grid, generating revenue for the user through feed-in tariffs. Conversely, in “Mode V,” solar output is low, and the charging demand is high. The battery discharges at its maximum rate, but even this is insufficient. The system then seamlessly connects to the grid to purchase the necessary power, ensuring the EVs are charged without interruption. The other modes represent transitional states, such as when the battery is charging at a constant voltage or discharging to support the load without needing grid power.
The mode-switching logic is critical. The system continuously monitors the power output from the PV system and the power required by the EVs. If solar power exceeds demand, the battery is instructed to charge. If solar power is less than demand, the battery is instructed to discharge. The system then checks the battery’s SOC to determine if it has reached its safe upper or lower limits, triggering a mode change if necessary. This creates a dynamic, self-regulating ecosystem that prioritizes the use of free solar energy, uses the battery to store excess and supply power when needed, and only resorts to the grid as a last resort for purchasing power or as a sink for surplus generation.
Ensuring the stability of the DC bus voltage is paramount for the safe and reliable operation of the entire microgrid. Voltage fluctuations can damage connected equipment and disrupt charging. To address this, the researchers divided the allowable voltage range into four bands: A, B, C, and D. Under normal conditions, the voltage is maintained near its nominal value of 750 V. If the voltage begins to rise into Band B—indicating an energy surplus—the PV system continues to operate at maximum power, while the battery switches to a constant-voltage charging mode to absorb the excess energy and pull the voltage back down. If the voltage continues to climb into the higher Band A, signaling that the battery is charging at its maximum capacity, the system hands over voltage control to the grid-side converter, which acts as a stabilizer to bring the voltage back to normal.
Similarly, if the voltage drops into Band C due to high demand, the battery discharges in constant-voltage mode to support the bus. If the voltage falls further into the lower Band D, indicating the battery is discharging at its maximum rate, the grid-side converter again takes over, injecting power to restore the voltage. This hierarchical voltage control strategy acts as a robust safety net, ensuring the system remains stable even during sudden changes in generation or load.
To enhance the performance of the PV system itself, the team implemented an improved version of the Maximum Power Point Tracking (MPPT) algorithm. MPPT is essential for extracting the maximum possible power from solar panels, which varies with sunlight intensity and temperature. The traditional “perturb and observe” method uses a fixed step size to search for the optimal operating point. This can be slow at first and causes the power output to oscillate around the maximum point, reducing efficiency.
The researchers’ modified algorithm uses a variable step size. It starts with a large step to quickly move towards the maximum power point, drastically reducing the initial search time. As it gets closer, the step size is progressively reduced, allowing for a much more precise and stable lock on the maximum power. Simulation results showed that this improvement increased the tracking speed by 18.3% and reduced power fluctuations by 35.4%, making the system more responsive to changing weather conditions and more efficient overall.
The feasibility and effectiveness of this coordinated control strategy were rigorously tested. The team first built a detailed simulation model in MATLAB/Simulink, running 24-hour scenarios for five distinct weather types: sunny, cloudy, rainy, overcast, and snowy. The results were impressive. In all cases, the DC bus voltage remained stable, with fluctuations kept below 5%, and the system smoothly transitioned between the six operational modes as predicted. On a sunny day, the system spent significant time in Mode II, selling excess solar power to the grid. On a snowy day, with heavy snow covering the panels, the system relied heavily on the grid in Mode V, but the control logic still functioned flawlessly.
To move beyond simulation and validate the strategy in a real-world context, the researchers conducted hardware-in-the-loop (HIL) experiments using a Yuankuan MT6020 real-time simulator. This advanced platform allows physical control hardware to interact with a virtual model of the power system, providing a highly accurate test environment. The experiments focused on the critical moments when the system switches between modes—such as when the sun comes out after a cloud, or when a new EV begins charging.
The HIL results confirmed the simulation findings. During a transition from Mode IV to Mode V, where the system connects to the grid to meet a sudden increase in demand, the DC bus voltage experienced a temporary dip of 7.53% but recovered to its nominal value within 0.25 seconds. All other transitions showed similar resilience, with voltage deviations quickly corrected. This demonstrates that the control strategy is not just a theoretical concept but a practical, robust solution capable of handling the fast dynamics of a real electrical system.
The final and most crucial part of the study was the economic analysis. The researchers calculated the daily charging cost under the five weather scenarios, factoring in the capital costs of the PV and battery systems, their expected lifespans (20 years for PV, 5 years for the battery), and the local electricity tariffs in Zhejiang, China. These tariffs include a peak rate of 0.588 yuan/kWh from 8:00 AM to 10:00 PM and a lower off-peak rate of 0.288 yuan/kWh for the rest of the day.
The results were compared against two benchmarks: the method proposed in a 2018 study by Li Lina et al. and the Whale Optimization Algorithm from a 2019 paper by Diab et al. The new strategy outperformed both in every scenario. On a rainy day, it reduced costs by 27.89% compared to Li’s method and by 26.03% compared to the Whale Optimization. Even on a sunny day, where the differences were smallest, it still achieved a 0.34% improvement. Most significantly, when compared to the cost of charging at a typical public charging station, which follows a different, often higher, tariff structure, the savings were enormous—up to 62.4%. This means that for many EV owners, investing in a smart solar-plus-storage system could pay for itself in a few years and then provide nearly free charging for the life of the vehicle.
This research represents a significant leap forward in the field of smart energy management. It moves beyond simple rule-based controls or computationally intensive optimization algorithms that require long calculation times. Instead, it offers a real-time, physics-based coordination strategy that is both highly effective and practical to implement. By intelligently managing the flow of energy between the sun, the battery, the grid, and the car, it transforms the EV from just a consumer of electricity into an active participant in a smarter, more resilient, and more economical energy ecosystem.
The implications of this work are far-reaching. For individual consumers, it offers a clear path to drastically lower their transportation costs. For utilities, widespread adoption of such systems could help to flatten the demand curve, reducing stress on the grid during peak hours. For policymakers, it provides a powerful tool to meet carbon reduction targets by maximizing the use of renewable energy and minimizing fossil fuel consumption.
While the study was conducted in a Chinese context, the principles are universally applicable. As the world continues its transition to electric mobility, strategies like this one will be essential for making EVs not just an environmentally sound choice, but a financially smart one as well. The work of Wang Chaoliang, Xiao Tao, Chen Songsong, Zhang Hongzhi, and Chen Ke, published in Electric Drive (DOI: 10.19457/j.1001-2095.dqcd25241), stands as a testament to the power of engineering innovation to solve real-world problems and drive sustainable progress.