New PWM Strategy Boosts EV Motor Efficiency and Reduces EMI
As the global automotive industry races toward electrification, one of the most critical challenges remains the optimization of electric drivetrains for maximum efficiency, reliability, and performance. While battery technology often dominates headlines, the underlying power electronics—particularly the inverter and its pulse width modulation (PWM) strategy—play a pivotal role in determining the real-world efficiency and longevity of electric vehicles (EVs). A groundbreaking study published in the Transactions of China Electrotechnical Society introduces a novel multi-mode modulation strategy based on Tri-State PWM (TSPWM), promising significant improvements in inverter efficiency while simultaneously reducing electromagnetic interference (EMI) and common-mode voltage (CMV)—a known culprit in premature motor bearing failure.
The research, led by Xia Yan from Shandong University of Technology, in collaboration with Sun Lipeng and Li Qiang from Weichai Power Company Limited and Li Junwei and Lu Haifeng from Tsinghua University, presents a comprehensive solution tailored specifically for the unique demands of EV motor drive systems. Unlike traditional approaches that rely on fixed-frequency modulation, this new strategy dynamically adapts to operating conditions, offering a smarter, more efficient way to manage power conversion.
The High-Frequency Dilemma in EV Drivetrains
Modern EVs demand high-performance electric motors capable of operating across a wide speed range, from low-speed city driving to high-speed highway cruising. To achieve smooth control and high dynamic response, inverters typically use high switching frequencies—often in the 10–20 kHz range. While this ensures excellent waveform quality and precise torque control, it comes at a cost: high switching losses in the power semiconductor devices, such as IGBTs or MOSFETs.
Switching losses are a major contributor to inverter inefficiency, especially at low motor speeds where output power is minimal but switching activity remains high. This imbalance results in poor system efficiency during common driving scenarios, ultimately reducing vehicle range. Moreover, high-frequency switching generates significant common-mode voltage, a high-frequency voltage that appears between the motor windings and ground. This CMV couples through parasitic capacitances in the motor and drivetrain, inducing shaft voltages that can break down the lubricating oil film in bearings, leading to electrical discharge machining (EDM) and premature bearing failure—a costly and reliability-threatening issue.
Traditional solutions, such as Space Vector PWM (SVPWM), while effective for control, exacerbate these problems. SVPWM frequently uses zero voltage vectors, which generate large CMV swings—up to half the DC bus voltage—and require frequent switching, increasing losses. Alternative strategies like Selective Harmonic Elimination PWM (SHEPWM) or synchronous modulation, often used in high-power traction systems, are computationally intensive and less suitable for the fast control loops required in EVs.
Introducing the TSPWM-Based Multi-Mode Strategy
The research team’s solution is a multi-faceted approach built around Tri-State PWM (TSPWM), a modulation technique that inherently reduces both switching losses and CMV. TSPWM operates by keeping one phase of the three-phase inverter clamped to either the positive or negative DC rail during each PWM cycle, effectively reducing the number of active switching devices from three to two. This simple yet powerful concept cuts the total number of switching events by approximately one-third compared to conventional SVPWM, directly translating to lower switching losses and higher inverter efficiency.
But the innovation doesn’t stop there. The team’s strategy introduces three key enhancements to make TSPWM even more effective in real-world EV applications.
First, the researchers developed a dynamic clamping mode selection algorithm. Instead of fixing the clamped phase based on the voltage vector sector, the system continuously monitors the phase currents and dynamically selects the phase with the highest current magnitude to remain clamped. Since the clamped phase only experiences conduction loss (which is relatively small) and no switching loss, aligning the clamp with the peak current phase maximizes loss reduction. This intelligent current-tracking approach ensures that the most loss-prone switching events—those occurring at high current—are minimized or eliminated.
Second, the team implemented a segmented variable carrier ratio modulation strategy. Rather than using a fixed switching frequency across the entire speed range, the system divides the operating envelope into multiple speed segments, each with an optimized carrier frequency. At low speeds, where high-frequency switching is unnecessary and wasteful, the carrier frequency is reduced to minimize switching losses. As motor speed increases, the carrier frequency is stepped up to maintain good current waveform quality and control performance. This adaptive approach strikes an optimal balance between efficiency and performance across the entire driving cycle.
The challenge with such a segmented strategy, however, is ensuring smooth transitions between different carrier frequencies. Abrupt changes in switching frequency can cause phase jumps in the reference voltage vector, leading to current and torque disturbances that manifest as jerks or vibrations in the vehicle. To solve this, the researchers developed a novel voltage vector phase compensation algorithm. By precisely calculating the angular displacement of the carrier cycle at the moment of switching, the algorithm applies a real-time phase correction to the reference voltage vector, ensuring a seamless and shock-free transition between modes. This level of control sophistication is critical for maintaining the smooth, refined driving experience expected in modern EVs.
Experimental Validation and Real-World Impact
The theoretical advantages of the proposed strategy were rigorously tested through both simulation and physical experimentation. The team used a permanent magnet synchronous motor (PMSM) commonly found in passenger EVs, controlled by a vector control system, and implemented the new TSPWM algorithm on a digital signal processor (DSP)-based controller.
The results were compelling. When compared to a conventional fixed-frequency SVPWM system operating at 10 kHz, the multi-mode TSPWM strategy demonstrated a significant improvement in inverter efficiency. The maximum efficiency increased by approximately 1%, while the average efficiency across the entire operating range improved by 5%. More importantly, the area of the efficiency map where the inverter operates at 95% efficiency or higher—the “high-efficiency zone”—expanded dramatically. For the traditional SVPWM, this high-efficiency zone covered only 33% of the total operating area. The variable-frequency SVPWM improved this to 68.3%, but the new TSPWM-based strategy pushed it to 73.1%, a substantial gain that directly translates to extended driving range.
The suppression of common-mode voltage was equally impressive. Measurements showed that the peak-to-peak CMV generated by the TSPWM algorithm was consistently one-third of that produced by SVPWM, regardless of the modulation index. This dramatic reduction not only mitigates the risk of bearing damage but also significantly lowers electromagnetic interference (EMI). Real-world tests on a trolleybus confirmed this: when using SVPWM, high-frequency EMI induced dangerous voltages (up to 280 V) on the controller housing, posing a safety risk. With TSPWM, these induced voltages were reduced to safe levels (below 36 V), demonstrating its effectiveness in creating a more robust and reliable electrical environment.
The smoothness of mode transitions was also validated experimentally. Without the phase compensation algorithm, switching the carrier frequency from 2 kHz to 10 kHz caused noticeable current oscillations and a 9% torque ripple—levels that would be perceptible and unpleasant to a driver. With the compensation algorithm active, the transition was seamless, with no visible current disturbance and torque ripple reduced to less than 2%. This level of control precision is essential for any advanced drivetrain strategy to be viable in production vehicles.
A Paradigm Shift in EV Power Electronics
The implications of this research extend far beyond a single efficiency number. It represents a paradigm shift in how engineers think about inverter control in EVs. Instead of treating the PWM strategy as a static, fixed component of the control system, this work demonstrates the power of a dynamic, adaptive approach that responds intelligently to changing operating conditions.
This is particularly relevant as the industry moves toward higher-performance vehicles with faster motors and more aggressive driving cycles. As motor speeds continue to climb into the 15,000–20,000 rpm range, the pressure on inverters to switch at ever-higher frequencies will only intensify. Strategies like the one proposed by Xia Yan and his colleagues offer a sustainable path forward, allowing engineers to maintain high performance without sacrificing efficiency or reliability.
Moreover, the benefits of reduced EMI and CMV are increasingly important as vehicles become more electrified and connected. High-frequency noise from the powertrain can interfere with sensitive sensors, communication systems, and advanced driver-assistance systems (ADAS). By minimizing this noise at its source, the TSPWM strategy contributes to a more stable and robust vehicle electrical architecture.
The computational simplicity of the algorithm is another key advantage. Unlike complex methods such as SHEPWM, which require extensive lookup tables and significant processing power, the TSPWM-based strategy is relatively lightweight and easy to implement on standard automotive microcontrollers. This makes it highly practical for mass production, where cost, reliability, and development time are critical factors.
Looking Ahead: From Research to Road
While the current study focused on a specific IGBT-based inverter, the principles of the multi-mode TSPWM strategy are broadly applicable. As the industry transitions to wide-bandgap semiconductors like silicon carbide (SiC) and gallium nitride (GaN), which can switch at even higher frequencies, the need for intelligent modulation strategies will only grow. These new devices offer the potential for even greater efficiency, but they also demand more sophisticated control to fully realize their benefits. The adaptive, loss-minimizing philosophy of this research provides a strong foundation for future innovations in power electronics.
In conclusion, the work by Xia Yan, Sun Lipeng, Li Junwei, Li Qiang, and Lu Haifeng represents a significant step forward in the optimization of EV motor drives. By combining the inherent advantages of TSPWM with dynamic current-based clamping, segmented carrier frequency control, and precise phase compensation, they have created a holistic solution that addresses multiple challenges simultaneously. The result is a more efficient, quieter, and more reliable drivetrain—one that brings the promise of longer range and lower ownership costs closer to reality. As automakers continue to refine their electric offerings, strategies like this will be essential in the quest for the ultimate electric driving experience.
Xia Yan, Sun Lipeng, Li Junwei, Li Qiang, Lu Haifeng, Transactions of China Electrotechnical Society, DOI: 10.19595/j.cnki.1000-6753.tces.222255