Torque ripple minimization for variable reluctance motors

Abstract

The design of current waveforms that minimize torque ripples in variable reluctance (VR) motors is addressed. The general problem of torque ripple reduction has been investigated by several researchers using various nonlinear models. The results have been design methods that require exhaustive measurements of motor characteristics. The approach presented here assumes magnetic linearity to derive an efficient ripple-reducing excitation that is much easier to calculate. A nonlinear optimal design problem is developed and then converted to a linear representation by a simple transformation. The transformed linear optimization problem is straightforwardly solved by standard numerical methods, and yields the optimal current waveform characteristics in the new variable. Finally, the optimal current waveform is found by applying an inverse transformation. The analytical result presented in this paper is supplemented by a design example that also addresses the performance tradeoff resulting from ignoring magnetic saturation to achieve simplifications in design.