Reliability-Based Robust Design Optimization With Kernel Density Estimation for Electric Power Steering Motor Considering Manufacturing Uncertainties

Abstract

Reliability-based robust design optimization (RBRDO) is a notable method to secure the quality and feasibility of performances from uncertainties. In this paper, geometric uncertainties of a stator sheet that can occur during a stamping process are considered as the uncertainty of controllable variables. Then, as the uncertainty of uncontrollable variables, skew angle due to inexact lamination is considered. No researches on the effect of this uncertainty have been performed. To analyze the effects of the uncertainties, the kernel density estimation (KDE) was employed to estimate the distribution because of its convenience, flexibility, and robustness. Then, RBRDO with the KDE is performed and the optimum results are compared with real data measured from manufactured motors.