Optimal Torque Sharing Function With Phase-Segmented Control for Switched Reluctance Motors

In this article, an improved torque sharing function (TSF) based on multiobjective optimization and phase-segmented control is presented in switched reluctance motor (SRM). First, the nondominated sorting genetic algorithm (NSGA) is adopted to globally optimize the full-phase turn-onangle, turn-offangle, and overlap angle, ensuring consistency of electromagnetic parameters across phases. A double-layer cooperative operation system (DCOS) with spatiotemporal symmetry is proposed for the windings of a six-phase motor, which is beneficial to the complementary superposition of torque vectors. The designed composite TSF incorporates nonlinear torque compensation to adjust torque distribution in commutation regions, mitigating torque deviations caused by inductance nonlinearity, current hysteresis, and residual currents. A bivariate cubic spline interpolation (BCSI) method is proposed to construct a globally smooth analytical torque model. The experimental results demonstrate that the proposed control method exhibits superior performance in aspects such as torque ripple reduction and copper loss minimization.