Design, Optimisation and Prototyping of Variable Reluctance Resolvers With Narrower Stator Excitation Teeth

Abstract

This paper aims to enhance the performance of the existing non-overlapping variable reluctance resolvers (VRRs) to obtain more accurate position signals by optimising the stator excitation teeth using genetic algorithm (GA). For this purpose, first, the principles of operation in non-overlapping VRRs are demonstrated using a magnetic equivalent circuit (MEC). Then, the MEC model is utilised to verify the principles of incorporating narrower excitation teeth in the non-overlapping model. The modified MEC provides a fast and sufficiently accurate model, so it is employed in the optimisation phase where GA is used to determine the optimal dimensions for the teeth. Following the MEC optimisation process, the proposed resolver is simulated and compared with a conventional equal teeth model using the finite element method (FEM). Finally, the proposed model with narrower excitation teeth is prototyped and tested. Results from MEC, FEM and the prototyped resolver confirm the effectiveness of the proposed model and validate the feasibility of using narrower excitation teeth to improve resolver's accuracy.