Harmonic Current Study Reveals New Path to Smoother EV Drives
In the relentless pursuit of quieter, more refined electric vehicles (EVs), a new study from Chongqing University of Technology is challenging conventional approaches to tackling one of the most persistent nuisances: drivetrain vibration and noise. As automakers push the boundaries of speed, power density, and integration in their e-axles, the complex interplay between electrical and mechanical components has created a sophisticated problem that cannot be solved by simply silencing individual parts. The culprit? A phenomenon known as harmonic current, an often-overlooked ripple in the electrical supply that can cascade through the entire powertrain, amplifying vibrations felt throughout the cabin.
This intricate challenge has now been dissected with unprecedented detail in a groundbreaking paper published in the Journal of Chongqing University of Technology (Natural Science). Led by Associate Professor Ge Shuaishuai from the university’s Key Laboratory of Advanced Manufacturing Technology for Automobile Parts, the research team has not only mapped how these electrical harmonics wreak havoc on system dynamics but has also pioneered an elegant, active solution that could redefine how engineers approach NVH (Noise, Vibration, and Harshness) in future EVs.
The core of the issue lies in the very nature of modern electric propulsion. Unlike an internal combustion engine with its inherent mechanical rhythms, an EV’s permanent magnet synchronous motor (PMSM) is driven by rapidly switching currents from an inverter. While this provides exceptional efficiency and control, it introduces high-frequency electrical distortions—harmonics—that are a byproduct of the pulse-width modulation process and non-linearities within the motor itself, such as magnetic saturation. Traditionally, NVH efforts have focused on passive solutions: adding dampers, optimizing gear tooth profiles, or improving mounting systems. However, these methods treat the symptoms, not the root cause, which is deeply embedded in the electromagnetic forces generated at the heart of the motor.
Ge Shuaishuai and his colleagues argue that a paradigm shift is necessary. “We need to move beyond viewing the motor and the gearbox as separate entities,” Ge explained. “They are fundamentally coupled. An electrical disturbance doesn’t just stay electrical; it translates into mechanical torque ripple, which then excites the gears, and the resulting mechanical vibrations can feedback and further distort the electrical currents. It’s a vicious cycle of electromechanical coupling.”
To prove this point, the team constructed a highly sophisticated dynamic model of a complete electric drive system. This wasn’t a simplified representation; it was a comprehensive simulation that integrated the real-world complexities often glossed over in earlier studies. The model incorporated the detailed structural characteristics of an 8-pole, 48-slot interior PMSM, including the non-linear effects of magnetic saturation and cross-coupling between the d-q axes. On the mechanical side, it accounted for the time-varying mesh stiffness of the gears, manufacturing errors, backlash, and bearing damping—all critical factors that contribute to gear whine and rattle.
By simulating the system under standard operating conditions—4,600 rpm and a load of 135 N·m—the researchers were able to observe the true nature of this coupling. Their findings were revealing. In the frequency spectrum of the electromagnetic torque, they didn’t just see the expected harmonics at 6fₑ and 12fₑ (where fₑ is the fundamental electrical frequency). They also detected distinct peaks corresponding to the first and second gear mesh frequencies (f_g1 and f_g2). Even more telling was the phase current, which showed not only the primary odd-order harmonics but also sidebands created by the modulation of the electrical frequency with the gear mesh frequencies. This was clear, empirical evidence that the mechanical vibrations of the gears were feeding back into the electrical domain, distorting the current waveform and creating a self-reinforcing loop of energy that manifests as audible noise and palpable vibration.
This discovery alone is significant, but the team went further. To isolate the impact of harmonic currents specifically, they conducted a series of controlled simulations where they injected artificial 5th, 7th, 11th, and 13th order harmonic currents into the system. The results were dramatic. With even a small injection of these harmonics, the peak-to-peak torque fluctuation surged from 82.06 N·m to 108.42 N·m—a 32% increase. The amplitude of the problematic 6fₑ and 12fₑ torque harmonics also ballooned. When larger harmonic currents were introduced, the torque ripple climbed to a staggering 134.57 N·m. More importantly, the study demonstrated that this increased electrical disturbance wasn’t confined to the motor. The heightened torque fluctuations were transmitted directly to the gears, causing a significant rise in the dynamic mesh force, particularly in the 6fₑ and 12fₑ components. This proved that harmonic currents are not a minor electrical curiosity; they are a primary driver of overall system vibration, especially as EVs operate at higher speeds where these frequencies become more prominent.
Armed with this deep understanding of the problem, the team turned their attention to a solution that is both counterintuitive and innovative: using harmonic currents to fight harmonic currents. The concept, known as active harmonic current injection, flips the traditional narrative on its head. Instead of trying to eliminate all harmonics, the strategy involves deliberately injecting specific, carefully calibrated harmonic currents into the motor’s control system to cancel out the unwanted torque ripple.
The physics behind this is rooted in the vector decomposition of electromagnetic torque. The total torque produced by a PMSM is a combination of a steady DC component and various AC harmonic components. These harmonic torques arise from interactions between harmonic currents, harmonic flux linkages, and cogging effects. By injecting a secondary set of harmonic currents with precisely tuned amplitude and phase, it is possible to create an opposing harmonic torque that destructively interferes with the original, problematic ripple.
“Think of it like noise-canceling headphones,” Ge said. “The headphones don’t just block sound; they generate a ‘negative’ sound wave to cancel the incoming noise. We’re doing the same thing with torque. We’re generating a ‘negative’ torque ripple to cancel the one caused by the motor’s inherent imperfections and external disturbances.”
Implementing this required a meticulous optimization process. The team identified the 6fₑ and 12fₑ harmonics as the primary contributors to objectionable vibration. They then treated the amplitude and phase of the 6th and 12th order harmonic currents in the d-axis and q-axis as variables in a massive parameter space. Using an offline optimization algorithm, they systematically searched for the exact combination of these four parameters (d-axis and q-axis amplitude and phase for each harmonic order) that would minimize the peak-to-peak torque ripple. This involved running thousands of simulations to map out the performance landscape, ultimately converging on a single, optimal set of values.
The final piece of the puzzle was the control architecture. To execute this strategy in real-time, the team designed a control system built around Proportional-Integral-Resonant (PIR) controllers for the d-axis and q-axis current loops. PIR controllers are ideal for this application because they can provide infinite gain at a specific resonant frequency, allowing them to track and inject the desired harmonic currents with exceptional accuracy, even in the presence of system variations.
The simulation results of this active control strategy were nothing short of impressive. When the optimized 6th and 12th order harmonic currents were injected at the 0.2-second mark, the effect was immediate. The motor’s torque ripple plummeted from 56.65 N·m to 23.49 N·m—a reduction of 58.5%. The amplitudes of the 6fₑ and 12fₑ harmonic components in the torque spectrum were slashed by more than 70%. Crucially, this benefit extended far beyond the motor. The output torque of the entire system saw its ripple decrease from 82.06 N·m to 47.65 N·m, a 41.2% improvement. Most significantly, the dynamic mesh forces in the gear train showed substantial reductions. The 6fₑ and 12fₑ components in the first-stage gear force were reduced by roughly 70%, while the dominant 6fₑ component in the second-stage gear force dropped by over 60%.
These results validate a powerful new approach to EV NVH. It moves the focus from passive, after-the-fact fixes to active, intelligent control that addresses the source of the problem. For automotive engineers, this means a potential path to achieving smoother, quieter driving experiences without resorting to heavier, more expensive mechanical solutions. It allows for the design of lighter, more compact gearboxes, knowing that a significant portion of the vibrational energy can be actively canceled at the source.
Moreover, the methodology has broader implications. The comprehensive modeling framework developed by Ge and his team provides a powerful tool for analyzing any integrated e-drive system. It can be used to predict vibration issues early in the design phase, evaluate the impact of different control strategies, or assess the sensitivity of a system to various non-linearities. This predictive capability is invaluable in an industry where development cycles are short and the cost of late-stage changes is prohibitive.
While the study is currently based on simulation, the path to real-world implementation is clear. The computational requirements for the offline optimization are manageable, and the PIR-based control strategy is well-suited for modern automotive microcontrollers. The next step will be experimental validation on a physical test bench, a challenge the team is already preparing for.
The work also highlights the importance of a holistic, multi-physics perspective in automotive engineering. As vehicles become increasingly electrified and software-defined, the boundaries between disciplines blur. Success will belong to those who can understand and master the complex couplings between electricity, magnetism, mechanics, and control theory. Ge Shuaishuai’s research is a prime example of this interdisciplinary thinking, offering a sophisticated yet practical solution to a problem that stands in the way of the ultimate EV experience: silent, seamless propulsion.
In conclusion, this research transcends a simple technical fix. It represents a fundamental shift in how we think about the electric drive. It shows that what was once considered an undesirable byproduct—harmonic current—can be transformed into a powerful tool for refinement. By embracing the complexity of electromechanical coupling rather than fighting against it, engineers can unlock a new level of performance and comfort, paving the way for a generation of EVs that are not just efficient, but truly serene.
Ge Shuaishuai, Zhao Jiayin, Zhang Zhigang, Guo Dong, Shi Xiaohui, Journal of Chongqing University of Technology (Natural Science), doi:10.3969/j.issn.1674-8425(z).2024.08.003