Position detection of BLDC rotor based on adaptive wavelet neural network

Brushless DC （BLDC） motor is a complex nonlinear system, of which some parameters will also change during operation. Therefore, obtaining accurate rotor position directly through the line voltage becomes more difficult. So a new method is proposed in this paper which uses three line voltages as the input signal to identify the motor position based on ada- ptive wavelet neural network （WNN） and the differential evolution （DE） algorithm to optimize WNN structures, thus real- izing the improvement of accuracy, exactness of the communication signals and convergence speed of the rotor position identification. Finally, both simulations and experimental results show that the proposed method has high accuracy of recognizing rotor position and strong orientation ability.