Genetic algorithm optimized on-line neuro-tuned robust position control of BLDC motor

In various control strategies of Brushless (BLDC) motor, PID controllers are still used due to their simplicity and ease of design. Unfortunately, PID controllers are not robust and their performance deteriorate when the operating conditions change due to the effect of external disturbances, load changes and parameter variations of the motor. A Robust PID controller which is optimized by Genetic algorithm and on-line tuned by Neural Network is proposed in this paper for position control of BLDC drive system. To optimize the controller performance due to changes in inertia and friction under dynamic load variation, estimation of inertia and friction at different load levels is done. The effectiveness of the controller is tested for set-point tracking and random changes in load torque. The results are compared with conventional tuning methods.