An optimal multi-objective control architecture of PMSM drives

Smarter Brains for Electric Car Motors

Electric cars rely on compact, powerful motors that must respond smoothly every time you press the accelerator or climb a hill. This paper explores a smarter way to control one of the most popular motor types in electric vehicles, aiming to make drives smoother, more efficient, and more reliable even when conditions change or parts age over time.

Why Motor Control Needs an Upgrade

Modern electric cars often use permanent magnet synchronous motors, which pack a lot of torque and efficiency into a small space. However, getting the best performance out of these machines is not simple. Traditional control methods, such as classic feedback controllers or fuzzy logic schemes, struggle when the motor behaves in a nonlinear way or when its internal properties drift with temperature, aging, or changing driving conditions. Other methods, like direct torque control, react quickly but cause large ripples in torque and current, which can translate into vibration, noise, and wasted energy. As electric vehicles spread and must cope with steep hills, frequent stop‑and‑go traffic, and varying loads, engineers need a control approach that is both fast and gentle on the hardware.

A Look Inside the New Control Method

The authors build on a strategy called model predictive control, which works by mathematically predicting how the motor will behave a short time into the future. At each instant, the controller evaluates possible actions and chooses the one that is expected to perform best according to a cost measure. In this study, that cost measure is “multi‑objective”: it balances several goals at once, such as keeping the motor current within safe limits, holding the supply voltage steady, and reducing power lost in the switching devices that drive the motor. A key innovation is a simplified “step‑ahead” model of the motor currents in a special rotating reference frame tied to the rotor. This makes the predictions fast enough to run at high sampling rates while still capturing the essential physics of torque production.

Making Fewer, Smarter Choices

One of the main challenges of model predictive control in power electronics is heavy computation. Every tiny time step, the controller could in principle test all possible switching combinations of the inverter that feeds the motor. The authors cut this burden by designing a four‑sector voltage selection scheme that considers only a reduced set of candidate voltage vectors, chosen based on the current error. A special nonlinear term in the cost function automatically excludes options that would drive the current beyond its safe peak, so the controller never seriously “overdrives” the motor. At the same time, a Lyapunov‑style energy measure is built into the objective, which mathematically guarantees that the system’s energy‑like quantity decreases over time, giving a firm foundation for stability.

Handling Real‑World Changes and Faults

The proposed controller is also designed with practical electric vehicles in mind. It regulates the voltage on the DC link—the internal power bus feeding the inverter—which lets it respond to sudden changes in load torque or road slope without losing control. Instead of relying on a physical speed sensor, which adds cost and complexity, the scheme uses current‑based information and a compact capacitor arrangement. Through detailed simulations in MATLAB/Simulink, the authors test extreme cases where the motor’s resistance and inductance are deliberately varied by as much as 50–150% of their rated values, mimicking heating, aging, and magnetic saturation. Even under these harsh conditions, the motor currents remain close to their desired values, torque settles quickly after disturbances, and voltage stays nearly constant.

What the Results Mean for Drivers

In simple terms, this work shows how a carefully designed “thinking” controller can keep an electric car’s motor running smoothly and efficiently even as the vehicle encounters hills, sudden accelerations, and long‑term wear. By predicting the motor’s behavior and weighing several goals at once—smooth torque, safe currents, and low switching losses—the proposed scheme reduces ripples in current, keeps total distortion below 5%, and cuts unnecessary switching. This combination promises quieter operation, better energy use, and greater robustness over the life of the vehicle. While the study is based on simulations, it lays a strong foundation for future experiments in real electric vehicles, where such intelligent control could ultimately help extend driving range and protect valuable battery and power‑electronics components.

Citation: Mohapatra, B.K., Sharma, V., Bhowmik, P. et al. An optimal multi-objective control architecture of PMSM drives. Sci Rep 16, 11289 (2026). https://doi.org/10.1038/s41598-026-38815-x

Keywords: electric vehicles, motor control, predictive control, permanent magnet motor, energy efficiency