Efficient dynamic performance of brushless DC motor using soft computing approaches

Abstract

A novel attempt to employ moth swarm algorithm (MSA) to generate the optimal gains of a proportional–integral (PI) speed controller of brushless DC (BLDC) motor is addressed to assure its satisfactory dynamic performance. For torque ripples minimization, a dual-loop speed controller is adapted. The agreed objective function is formulated to minimize the integral time absolute speed error (ITAE) subjects to set of constraints. The effectiveness of the MSA is verified through many test cases along with the detailed comparisons to those obtained by well known genetic algorithm and particle swarm optimization. At this stage, the numerical results of the MSA are used to train and test an artificial neural network which shall be used as an adaptive controller to give the optimal PI gains under different operating conditions. At final stage, the performance of the BLDC motor powered from photovoltaic (PV)–battery hybrid system with the proposed controller is demonstrated. A Landsman converter is controlled by an incremental conductance technique to maximize the extracted PV array power. A bidirectional converter is used to control battery charging/discharging states. Various demonstrated case studies indicate that the MSA is effective in generating the optimal gains of the PI controller.