Multi-objective PSO/GA optimization control strategies for energy efficient PMDC motor drives

Abstract

The paper presents the application of genetic algorithm (GA) and particle swarm optimization (PSO) search and optimization techniques for online tuning of error-driven self-adjusting multi-loop dynamic speed regulators for large industrial permanent magnet DC (PMDC) motor drives. Novel three multi-loop dynamic error regulator include tan-sigmoid controller, multi-zone controller, and incremental self-regulating controllers are developed by First Author and validate by Second Author. The proposed dynamic multi-loop control strategies are based on time-discalced and decoupled regulation loops with assigned loop weightings and dynamic loop delays. The three novel dynamic and efficient control schemes utilize speed, current, dynamic momentum excursion error, and limited current ripple errors as the main inputs to vary the firing delay angle α of the six-pulse controlled Thyristor rectifier. Multi-objective GA (MOGA) and multi-objective PSO (MOPSO) are used for selection of the proposed controller's gains to ensure high performance efficient PMDC motor drives and effective robust dynamic tracking of different speed reference trajectories. The optimization process is based on minimizing the system control total error, the steady state error, settling time, rising time, and maximum overshoot. Copyright © 2011 John Wiley & Sons, Ltd.