Implementation and study of saliency tracking-based sensorless control methods for a permanent magnet synchronous motor in an e-bike application

Details

Supervisors

Dehez, Bruno

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil électromécanicien, à finalité spécialisée: mécatronique
Master [120] : ingénieur civil électromécanicien, à finalité spécialisée: énergie

Abstract

This master’s thesis explores the implementation of sensorless control methods for a permanent magnet synchronous motor (PMSM) in an e-bike application. Specifically, the study focuses on saliency tracking-based techniques, particularly high-frequency signal injection (HFSI), to achieve reliable and efficient low-speed rotor position estimation without relying on physical sensors. The research assesses the feasibility of these methods within the constraints set by E2Drives, considering factors such as precision, robustness, power consumption, and acoustic noise. The thesis begins with a comprehensive review of existing low-speed sensorless control methods, providing a rationale for selecting square-wave injection for further implementation. A controller model is then developed within a Simulink framework, integrating the sensorless control algorithm along with other control modes necessary for e-bike operation. The model is implemented and tested on a dedicated test bench, focusing on performance evaluation under various conditions, including error signal analysis, steady-state and dynamic performance, and power consumption. Key findings indicate that while the sensorless control method demonstrates strong performance in speed tracking and dynamic capabilities, challenges remain in maintaining stability under high load and speed conditions. The noise reduction techniques are effective in mitigating acoustic disturbances, but they introduce some uncertainties in position tracking. Moreover, the dual pulse method for initial position estimation performs well, albeit with occasional failures due to current measurement limitations. While the study successfully addresses many of the defined objectives, further refinements in algorithm design and hardware optimization are recommended to enhance stability and extend the operational limits of the control system. The findings contribute valuable insights for future developments in sensorless control methods for e-bike applications, highlighting the need for continued research to meet the demands of real-world conditions.