New Control Strategy Boosts Stability of Off-Grid Highway Microgrids with Electric-Hydrogen Integration
In the rapidly evolving landscape of sustainable transportation infrastructure, the integration of renewable energy sources with electric vehicle (EV) charging networks presents both immense opportunity and formidable technical challenges. A groundbreaking control strategy developed by researchers at North China Electric Power University and Beijing Jiaotong University offers a sophisticated solution to stabilize off-grid highway microgrids, particularly those incorporating electric-hydrogen coupling systems. This innovation addresses the critical issue of voltage instability caused by the unpredictable nature of EV fast-charging demands and intermittent renewable generation, while simultaneously optimizing the lifespan of battery storage units.
The research, published in the journal Electric Machines and Control Applications (DOI: 10.12177/emca.2024.033), introduces a novel multi-source coordinated control strategy based on bus voltage deviation and State of Charge (SOC) management. Led by ZHAO Yihui, CHENG Peng, TIAN Xinshou, and JIA Limin, the team’s work provides a robust framework for ensuring reliable power supply in remote highway service areas, which are increasingly becoming vital nodes in the national electric mobility network. Their approach not only enhances system resilience but also maximizes the utilization efficiency of renewable energy, paving the way for more sustainable and autonomous energy systems along major transportation corridors.
Highway service areas, unlike industrial parks or urban grids, face unique operational constraints. They typically lack connection to the main utility grid, relying instead on localized generation from solar photovoltaic (PV) arrays and, to a lesser extent, wind turbines. These facilities must support a diverse array of loads, including lighting, ventilation, communications, and most critically, EV charging stations. The challenge lies in the inherent unpredictability of EV charging behavior. Unlike residential or workplace charging, where users often plug in overnight or during extended periods, highway travelers demand rapid replenishment—often within half an hour—to continue their journeys. This results in sudden, high-power draw events that can destabilize the local microgrid, causing voltage sags or swells that risk damaging equipment and interrupting service.
Traditional solutions have relied heavily on battery energy storage systems (BESS) to absorb these transient power fluctuations. While batteries offer excellent response speed and high power density, they suffer from limited energy capacity and degradation when subjected to frequent deep discharge cycles or overcharging. In environments where renewable generation is intermittent—such as during cloudy days or low-wind periods—the BESS may be unable to sustainably balance supply and demand without depleting its charge or exceeding safe operating limits. This limitation has spurred interest in hybrid storage systems that combine batteries with hydrogen-based technologies, such as electrolyzers for hydrogen production and fuel cells for electricity generation.
Hydrogen storage offers several advantages over batteries alone. It possesses a much higher energy density, allowing for longer-duration energy storage without significant physical footprint increases. Hydrogen can also be stored indefinitely without self-discharge, making it ideal for seasonal or long-term energy buffering. However, hydrogen systems exhibit slower dynamic response times compared to batteries, which can delay their ability to react to sudden load changes. The key insight of the new control strategy is to intelligently coordinate these complementary technologies—not merely treating them as separate components, but as an integrated, adaptive system capable of responding dynamically to real-time conditions.
The core of the proposed methodology revolves around using the DC bus voltage as the primary coordination signal. Rather than relying on complex communication networks between distributed energy resources—a common feature in centralized control architectures—the system leverages the natural electrical characteristics of the microgrid itself. Voltage deviations from the nominal level (set at 1500V in this study) serve as direct indicators of power imbalance: a rising voltage suggests excess generation, while a falling voltage indicates insufficient supply. By mapping these voltage levels to predefined operational modes, the system can automatically adjust the behavior of each energy source and storage unit without requiring external commands.
Four distinct operational modes were defined based on voltage thresholds:
- Mode 1 (U_bus > 1575V): Power surplus condition. Renewable generators reduce output to avoid overvoltage; electrolyzers operate at maximum capacity to convert excess electricity into hydrogen; batteries enter droop control mode to stabilize voltage; fuel cells remain idle.
- Mode 2 (1425V < U_bus < 1575V): Balanced condition. Generators operate at maximum power point tracking (MPPT); both batteries and electrolyzers participate in voltage regulation via droop control; fuel cells remain idle.
- Mode 3 (1350V < U_bus < 1425V): Power deficit condition. Generators continue MPPT operation; batteries and fuel cells share the responsibility of supplying additional power through droop control; electrolyzers shut down to conserve energy.
- Mode 4 (U_bus < 1350V): Severe power deficit. Fuel cells operate at maximum output; batteries provide support; if voltage continues to drop below the minimum threshold, non-critical loads are shed according to pre-established priority schemes.
This hierarchical, voltage-based switching mechanism ensures that the system responds appropriately to varying degrees of stress without overwhelming any single component. Crucially, it minimizes the need for frequent controller mode switches, which can introduce transients and degrade power quality. Instead, the system relies on continuous, smooth adjustments governed by droop characteristics—essentially mimicking the frequency response behavior found in AC grids, but adapted for DC applications.
However, the true innovation lies in the introduction of an adaptive droop control mechanism that incorporates SOC feedback. Conventional droop control uses fixed coefficients, meaning that regardless of how full or depleted the batteries are, they will always respond to voltage deviations in the same manner. This rigidity can lead to premature aging of the battery pack, especially if it is repeatedly forced into deep discharge or overcharge states to maintain voltage stability. Recognizing this, the research team designed a dynamic adjustment algorithm that modifies the droop coefficients of both the battery system and the hydrogen subsystem based on the current SOC levels.
When the battery SOC is high, indicating ample stored energy, the droop coefficient for the battery is increased during charging phases (when bus voltage exceeds nominal). This effectively reduces the rate at which the battery accepts charge, preventing it from reaching its upper limit too quickly. Simultaneously, the droop coefficient for the electrolyzer is also increased, encouraging it to absorb more excess power and produce hydrogen instead. Conversely, when the battery SOC is low, the droop coefficient for the battery is decreased during discharging phases (when bus voltage falls below nominal), allowing it to deliver more power initially while gradually reducing its contribution as SOC declines. To compensate for this reduced battery output, the droop coefficient for the fuel cell is increased, prompting it to ramp up its power generation more aggressively.
This intelligent redistribution of power responsibilities among the various energy sources serves two critical purposes. First, it protects the battery from excessive cycling and extreme SOC excursions, thereby extending its operational life and reducing maintenance costs. Second, it mitigates the voltage deviations that would otherwise occur if the battery alone had to handle all imbalances, resulting in improved overall power quality and system reliability. The researchers further refined this approach by introducing an optimization factor K, which allows fine-tuning of the convergence rate toward target SOC values. Too large a K value slows down the adjustment process, potentially leaving the system vulnerable to prolonged imbalances. Too small a K value accelerates convergence but risks inducing oscillations or instability. Through simulation studies, the team determined that a K value of 0.1 provided an optimal balance between responsiveness and stability.
To validate their theoretical framework, the researchers constructed a detailed simulation model using MATLAB/Simulink, replicating the architecture of a typical off-grid highway microgrid. The simulated system included a 300kW wind turbine, a 2.2MW PV array, an 880kW electrolyzer, a 900kW fuel cell, and multiple battery banks with capacities of 300Ah each. Load profiles were carefully crafted to reflect realistic EV charging patterns observed at highway service areas, characterized by sharp peaks around midday and evening hours corresponding to peak travel times. Four distinct environmental scenarios were tested: high wind/high sun, high wind/low sun, low wind/low sun, and low wind/high sun. Each scenario was paired with varying levels of constant power and resistive loads to simulate different traffic volumes and usage patterns.
Simulation results demonstrated clear advantages of the proposed adaptive control strategy over conventional fixed-droop methods. In high-SOC scenarios, where batteries were already near full charge, the adaptive controller successfully suppressed unnecessary battery charging while increasing hydrogen production, leading to a 15% reduction in voltage deviation compared to baseline performance. Similarly, in low-SOC scenarios, where batteries needed protection from deep discharge, the controller enabled fuel cells to assume a greater share of the load, reducing voltage sag by approximately 20%. Moreover, the adaptive strategy significantly improved SOC balancing across multiple battery units, reducing inter-battery SOC differences from 5% under fixed control to less than 4.5% after 240 seconds of operation.
These findings underscore the practical relevance of the research. For highway operators and energy providers, implementing such a control strategy could translate into tangible benefits: reduced downtime due to voltage instability, lower maintenance costs associated with battery replacement, increased utilization of locally generated renewable energy, and enhanced customer satisfaction through uninterrupted EV charging services. From a broader perspective, this work contributes to the development of resilient, decentralized energy systems that can operate autonomously even in remote locations, supporting the transition toward carbon-neutral transportation infrastructure.
It is worth noting that while the current implementation focuses on isolated microgrids, the principles underlying this control strategy are scalable and adaptable. With minor modifications, similar approaches could be applied to larger microgrid clusters, community energy systems, or even islanded utility grids. Furthermore, the modular nature of the design allows for easy integration of additional energy sources or storage technologies as they become available, ensuring future-proofing against technological advancements.
Looking ahead, the research team plans to extend their work in several directions. One area of focus is enhancing the predictive capabilities of the controller by incorporating machine learning algorithms that can forecast EV arrival patterns and renewable generation trends based on historical data. Another avenue involves exploring the economic implications of the strategy, including cost-benefit analyses comparing capital expenditures for hybrid storage systems versus traditional battery-only configurations. Additionally, field trials in real-world highway service areas are being planned to assess the long-term durability and operational effectiveness of the system under actual environmental and usage conditions.
In conclusion, the multi-source coordinated control strategy developed by ZHAO Yihui, CHENG Peng, TIAN Xinshou, and JIA Limin represents a significant step forward in the engineering of sustainable, self-sufficient energy systems for transportation infrastructure. By harmonizing the strengths of batteries and hydrogen technologies through intelligent, adaptive control, the researchers have created a blueprint for stable, efficient, and environmentally responsible microgrids that can meet the growing demands of electrified mobility. As nations worldwide strive to decarbonize their transport sectors, innovations like this will play a pivotal role in enabling seamless, reliable, and scalable adoption of electric vehicles—even in the most challenging and remote settings.
ZHAO Yihui, CHENG Peng, TIAN Xinshou, JIA Limin, North China Electric Power University, Beijing Jiaotong University, Electric Machines and Control Applications, DOI: 10.12177/emca.2024.033