A Neural Network-PSO Based Control for Brushless DC Motors for Minimizing Commutation Torque Ripple

This paper presents the method of reducing torque ripple of brushless DC (BLDC) motor. The commutation torque ripple is reduced by control of the DC link voltage during the commutation time. The magnitude of voltage and commutation time is estimated by a neural network and optimized with an optimization method named particle swarm optimization (PSO) algorithm analysis. The goal of optimization is to minimize the error between the command torque and real torque and doesn’t need knowledge of the conduction interval of the three phases. It adaptively adjusts the DC link voltage in commutation duration so that commutation torque ripple is effectively reduced. In this paper, the performance of the proposed brushless DC (BLDC) control is compared with that of conventional BLDC drives without input voltage control.

Keywords:

BLDC machines , Commutation , Optimized input voltage , Torque ripple