Low-Speed Cutoff Point Detection During Regenerative Braking in Electrified Vehicles
A sensorless, model-based method that lowers the regenerative braking cutoff speed to recover more energy and extend vehicle range.
A key limitation of regenerative braking in electric vehicles is its inability to recover energy at low speeds, where the traction motor develops insufficient back-electromotive force. Below a critical speed threshold—the low-speed cutoff point—energy is drawn from the battery rather than returned to it, as the system works to overcome electrical losses in the motor drive. Dynamic threshold-detection methods are known to deliver higher braking efficiencies than fixed-point approaches, but conventional implementations rely on sensing the direction of battery current, which introduces challenges related to sensor accuracy, offset, filtering, and delay. This study instead develops a model-based approach that determines the dynamic low-speed cutoff point analytically. By pairing this with a loss-minimization controller, we achieve a lower cutoff point that extends the effective braking window and increases the energy recovered. We further examine the resulting improvements in vehicle range and overall braking performance under realistic driving conditions.
Systematic Component Sizing and Control for EV Traction Drives
An impedance-based framework for sizing passive components that avoids over-design while preserving system stability.
The sizing of passive components in EV drive systems is a decisive design choice, shaping cost, power density, performance, and efficiency alike. Component sizing has traditionally relied on empirical design rules applied to worst-case scenarios, an approach that often yields over-sized and inefficient designs. This work develops a systematic sizing methodology that makes full use of the available hardware without compromising system stability. Using impedance-based stability analysis, we identify the optimal component values that balance robustness against efficient hardware utilization.
Single-Sensor Control Techniques for EV Traction Drives
Reconstructing phase currents from a single DC-link sensor using parameter-free observers with frequency-adaptive pole placement.
Single-sensor control offers a low-cost route to implementing multi-phase AC drives. Reconstructing the individual phase currents from a single DC-link current sensor typically requires model-based observers that depend on machine load parameters. As an alternative, we explore sinusoidal curve-fitting observers, which remove this dependence entirely. In such observers, the poles governing the error dynamics set the rate at which estimation errors are corrected: fixed-gain designs perform well at a nominal frequency but degrade under variable-frequency operation. This work investigates frequency-adaptive pole-placement techniques that minimize current-reconstruction error while ensuring stable tracking of the reference currents across a wide range of operating frequencies.
High-Performance, Loss-Minimizing Control Techniques for Motor Drives
Control strategies that break the usual trade-off between fast dynamic response and low energy loss across mixed city–highway drive cycles.
This research develops control techniques that sharpen dynamic response while reducing energy losses across demanding drive cycles that span both city- and highway-like driving conditions. By studying variable-flux operation together with careful choices of sampling time for methods such as deadbeat control, we aim to break the traditional trade-offs between dynamic performance and efficiency—addressing limitations that constrain conventional implementations.
