New Control Strategy Enhances Range Extender Efficiency in Electric Vehicles
The rapid evolution of electric vehicles (EVs) has placed increasing demands on auxiliary power systems, particularly range extenders used in extended-range electric vehicles (EREVs). These systems, typically composed of an internal combustion engine (ICE) coupled with a generator, play a pivotal role in extending driving range and maintaining battery charge levels without direct reliance on charging infrastructure. However, despite their utility, range extenders have long faced challenges in dynamic power control, including delayed response times, inaccurate power output, and transient overshoots—issues that compromise fuel efficiency, emissions performance, and overall drivability. A recent breakthrough by a research team from Zhejiang University offers a promising solution to these longstanding problems.
In a study published in the Chinese Internal Combustion Engine Engineering journal, Yao Dongwei, Shen Junhao, Wu Feng, and Lu Xinwei from the Power Machinery and Vehicular Engineering Institute at Zhejiang University, along with collaborators from the Key Laboratory of Smart Thermal Management Science & Technology for Vehicles of Zhejiang Province, have introduced a novel dynamic control strategy designed to significantly improve the performance of range extender systems. Their approach, which centers on a power-speed dual-loop control framework, addresses the core limitations of conventional control methods by integrating real-time feedback mechanisms and dynamic coordination between engine torque and generator speed.
The research team’s work emerges at a critical juncture in the global transition toward electrified transportation. While battery electric vehicles (BEVs) continue to dominate headlines, EREVs have carved out a unique niche by combining the benefits of full electric driving with the range assurance of an onboard generator. This hybrid configuration eliminates range anxiety—a major barrier to EV adoption—while maintaining high levels of energy efficiency and low emissions during typical urban driving conditions. In China, where the government has prioritized new energy vehicles as a strategic industry, EREVs are increasingly seen as a pragmatic bridge between traditional internal combustion vehicles and a fully electric future.
However, the effectiveness of an EREV hinges on the seamless integration and control of its subsystems. Among these, the range extender must respond rapidly and accurately to power requests from the vehicle’s energy management system. Traditional control strategies often rely on open-loop torque models for the engine and independent speed control for the generator. While conceptually straightforward, this decoupled approach fails to account for the inherent mismatch in response dynamics between the mechanical inertia of the ICE and the near-instantaneous electromagnetic response of the generator. As a result, when the vehicle demands a sudden change in power output—such as during acceleration or regenerative braking transitions—the system frequently exhibits lag, overshoot, or even reverse power fluctuations, where the generator momentarily draws power instead of supplying it.
Yao Dongwei and his team recognized that these transient instabilities stem from two primary sources: first, the absence of real-time torque feedback in ICE control, and second, the lack of coordination between torque and speed commands during dynamic load changes. To address these issues, they developed a dual-loop control architecture that fundamentally rethinks how power generation is managed in a range extender system.
At the heart of their strategy is the concept of power-speed decoupling based on an optimized efficiency curve. Rather than treating the range extender as a simple power source, the researchers mapped out a comprehensive efficient operating curve that defines the ideal combination of engine torque and generator speed for each target power output. This curve was derived from extensive testing of both the engine’s fuel consumption characteristics and the generator’s electrical efficiency across various load points. By referencing this pre-calibrated map, the control system can instantly determine the most efficient mechanical power input required to achieve a given electrical output, factoring in losses from the generator and its power electronics.
Once the target mechanical power is estimated, it is decomposed into a target torque and a target speed. These values serve as the initial setpoints for the engine and generator controllers, respectively. However, the innovation lies not in this decomposition itself, but in the subsequent refinement of these setpoints through closed-loop feedback mechanisms.
The first layer of correction involves a power feedback loop. Since internal combustion engines do not have direct torque sensors, their output torque is typically inferred from parameters such as air-fuel ratio, ignition timing, and intake manifold pressure. These models, while sophisticated, are prone to calibration errors and drift over time due to component aging or environmental changes. To compensate for these inaccuracies, the team implemented a proportional-integral (PI) controller that continuously monitors the difference between the commanded power and the actual power output measured at the generator terminals. This error signal is then used to adjust the target torque command in real time, effectively closing the loop on power delivery.
This feedback mechanism ensures that even if the engine’s actual torque deviates from its intended value, the system can correct the discrepancy before it leads to significant power errors. The result is a dramatic reduction in steady-state power deviation—so much so that in their experimental validation, the researchers reported a steady-state error of less than 0.01 kW across all tested conditions. This level of precision is remarkable, especially considering the complex, nonlinear behavior of combustion engines.
The second layer of the control strategy addresses the dynamic mismatch between torque and speed response. In conventional systems, when a power increase is requested, the generator rapidly adjusts its speed to match the new setpoint, often before the engine has had time to build up the necessary torque. This temporal misalignment causes the generator to momentarily reduce its electrical load to accommodate the speed change, leading to a drop in output power—a phenomenon known as reverse overshoot. Conversely, during a power reduction, the generator may briefly increase its load to slow down, causing a spike in power output.
To mitigate this issue, the researchers introduced a dynamic coordination mechanism that deliberately slows down the generator’s speed response using a first-order filtering process. By applying a time constant to the target speed command, they ensured that the generator’s speed changes occur at a rate more closely matched to the engine’s torque development. This synchronization prevents abrupt load shifts and allows both components to transition smoothly between operating points.
The effectiveness of this coordinated approach was demonstrated through a series of transient load tests conducted on a benchtop range extender system. The setup included a 1.5-liter naturally aspirated gasoline engine directly coupled to a water-cooled permanent magnet synchronous generator—the same configuration found in many production EREVs. The system was connected to a programmable DC power supply that simulated the vehicle’s high-voltage bus, allowing the researchers to precisely control and measure power flow.
During the experiments, the target power was stepped from 5 kW to 20 kW in 5-kW increments, then ramped back down to 5 kW. At each step, the system’s response was recorded, with particular attention paid to overshoot, response time, and tracking accuracy. The results were striking: the maximum positive overshoot was limited to just 1.37 kW, while the maximum negative (reverse) overshoot was 1.89 kW. The rise time for power increases was approximately 2.5 seconds, and the fall time for decreases was about 2.8 seconds—both figures representing a substantial improvement over conventional control methods.
More importantly, the operating points throughout the test remained closely aligned with the pre-defined efficient operating curve. This consistency ensures that the system operates at peak efficiency not only in steady-state conditions but also during transient events, which are common in real-world driving. The ability to maintain high efficiency during dynamic operation is a key advantage, as it directly translates into lower fuel consumption and reduced emissions.
One of the most compelling aspects of the study is the transparency with which the researchers documented the underlying causes of performance limitations. Through detailed analysis of torque and speed data, they showed that the uncorrected engine torque command differed from the actual torque by an average of 8.5 N·m across different load points. This discrepancy, while not uncommon in production engines, would typically result in noticeable power errors if left uncorrected. However, thanks to the feedback-based torque correction, these deviations were effectively compensated, resulting in near-perfect power tracking.
The implications of this research extend beyond the laboratory. For automakers developing EREVs, the proposed control strategy offers a practical and scalable solution to one of the most persistent challenges in range extender design. Unlike some advanced control techniques that require extensive hardware modifications or complex machine learning models, this approach relies on software-based enhancements that can be implemented within existing electronic control units (ECUs). The use of standard PI control and first-order filtering makes the strategy computationally efficient and robust, qualities that are essential for real-time automotive applications.
Moreover, the methodology is adaptable to different engine-generator pairings and can be fine-tuned based on specific vehicle requirements. For instance, in a luxury EREV where smoothness and quiet operation are paramount, the time constant for speed filtering could be increased to further dampen transients. In contrast, for a performance-oriented model, the response could be sharpened to prioritize agility over absolute smoothness.
The work also highlights the importance of system-level thinking in automotive engineering. Rather than optimizing individual components in isolation, the researchers took a holistic view of the range extender as an integrated power unit. This systems approach enabled them to identify and resolve interactions between subsystems that would otherwise go unnoticed. It serves as a reminder that in the pursuit of efficiency and performance, the whole is often greater than the sum of its parts.
From a broader industry perspective, this research underscores the ongoing relevance of internal combustion technology in the electrified era. While the long-term trajectory points toward full electrification, range extenders will likely remain a key component in the mobility ecosystem for years to come, particularly in regions with underdeveloped charging infrastructure or for commercial vehicles with high daily mileage requirements. Improving the efficiency and responsiveness of these systems not only enhances the competitiveness of EREVs but also contributes to the overall reduction of carbon emissions in the transportation sector.
In conclusion, the dynamic control strategy developed by Yao Dongwei, Shen Junhao, Wu Feng, and Lu Xinwei represents a significant advancement in the field of range extender technology. By combining power-speed decoupling, real-time power feedback, and dynamic torque-speed coordination, their approach effectively addresses the core challenges of transient power control. The experimental results demonstrate exceptional performance in terms of accuracy, response time, and efficiency, making the strategy a strong candidate for implementation in future production vehicles. As the automotive industry continues to navigate the complexities of electrification, innovations like this one will play a crucial role in shaping the next generation of sustainable transportation solutions.
Yao Dongwei, Shen Junhao, Wu Feng, Lu Xinwei, Zhejiang University, Chinese Internal Combustion Engine Engineering, DOI: 10.13949/j.cnki.nrjgc.2024.01.008