Optimization of Analytical TSFs for DC-Link Current Reduction in Switched Reluctance Motors

Abstract

The large dc-link current is a known issue in switched reluctance motor (SRM) drives, which often demand the use of a bulky dc-link capacitor. However, control techniques can be designed and optimized to lessen this issue. In this context, this article proposes the optimization of analytical torque sharing functions (TSFs) for dc-link current reduction in SRMs. Initially, the analytical TSFs are described, and the importance of adequate parameter selection is highlighted. Next, an optimization procedure based on the nondominated sorting genetic algorithm II is proposed to determine the optimal turn- on and overlap angles by solving a multiobjective optimization problem considering torque rms error and dc-link rms current as objectives to be minimized, something not previously reported in the literature. The pareto fronts for different operating conditions are presented, including both soft and hard chopping operation, as well as different sampling frequencies. Then, an approach for selecting a solution from within the pareto front is described, enabling the result from the pareto front that yielded the desired tradeoff between torque RMSE and dc-link current to be identified. Experimental results are provided to support the effectiveness of the proposal. A comparison between three different cases is shown, highlighting the tradeoff between objectives.