Artificial neural network based proportional plus integral plus derivative controller for a brushless DC position control system

Abstract

Brushless DC (BLDC) motors are useful in applications such as process control, robotics, industrial automation, aerospace, electric vehicles etc. due to such advantages as the elimination of rotor losses and magnetizing current. Wider usage of BLDC motors demands optimum position control for high efficiency, accuracy and reliability. For an accurate position control, the estimation of moment of inertia and friction coefficient of the motor with load is essential. This paper incorporates the computation of the moment of inertia and friction coefficient of a BLDC motor with load at different load settings and emphasizes that load has an appreciable effect on the dynamic performance of the system. To obtain the optimum position control, a proportional plus integral plus derivative (PID) controller is employed and tuned using the PARR method. Artificial neural networks are used for computing the moment of inertia and friction coefficient of a BLDC motor with load and PID controller parameters at various load settings. Simulation results of the position control system are obtained at different load settings. From the results, it is evident that the PID controlled position control system responds to the desired position with minimum rise time, settling time and peak overshoot and is in dependent of load settings.