Hall sensors are used to estimate angular velocity of brushless DC motors. However, the estimation performance deteriorates due to misalignment of Hall sensors and magnetic poles on the rotor. The proposed method uses a particle filter to robustly estimate angular velocity against the misalignment. Specifically, the estimation performance is improved by reflecting the distribution of the misalignment in likelihood design of particle filter. Furthermore, the distribution of the misalignment can be easily obtained through preliminary experiments. Experiments verify that the designed likelihood improves estimation performance of the particle filter.
Muto et al. (2025) studied this question.