Systematic Metamodel-Based Optimization Study of Synchronous Reluctance Machine Rotor Barrier Topologies

Abstract

Nowadays, due to the confidence in modeling tools and rapid product iteration, electric machine designers primarily rely on simulations. This approach reduces time and cost and is very useful when comparing different machine topologies. The prototype stage usually comes after the depletion of all simulation resources. When designing a synchronous reluctance machine, the first step is the selection of rotor barrier type. The literature provides several topologies but does not clearly state which one yields the best performance. The goal of this paper is to determine the best variant for a six-pole machine and the selected requirements using a metamodel-based optimization approach. Seven rotor topologies with different complexities were derived from circular, hyperbolic, and Zhukovsky barrier types (circular concentric, circular variable depth, hyperbolic with fixed eccentricity, hyperbolic with variable eccentricity, original Zhukovsky, modified Zhukovsky variable depth and modified Zhukovsky with equal barrier depth). The novelty of the proposed strategy is in the systematic and fair comparison of different rotor topologies. This approach significantly reduces the total optimization time from several weeks to a few days. Additionally, a novel modified Zhukovsky variable depth topology, which merges the best qualities of all considered variants, was developed. An identical optimization strategy was applied to all variants, and the final results prove that the barrier type substantially affects the final performance of the machine. The best results are achieved by the modified Zhukovsky variable depth topology. In relation to the worst (baseline) topology, the performance gain is 14.9% and the power factor is increased from 0.61 to 0.67. An additional study using different numbers of barrier layers (3, 4, and 5) was conducted to determine the best topology. The best results were achieved with the original four barrier layers.