IEEE Transactions on Magnetics · 2019 · 69 citations · 32 references
ArchitectureManufacturing EngineeringMechanical EngineeringDesign OptimizationMachine DesignOptimizationHybrid Vehicle SimulationOptimization Under UncertaintyResponse Surface MethodologyRobust Design OptimizationRobust OptimizationSurrogate-based OptimizationDeterministic OptimizationHigh-dimensional Surrogate ModelingHybrid Vehicle DesignMonte Carlo AnalysisParticle Swarm OptimizationPermanent Magnet Motor DesignRobust DesignSix Sigma Methodology
Industrial manufacturing is subject to uncertainties that can negatively affect product quality when deterministic optimization is used. The study aims to quantify uncertainties in the design process to achieve desired product quality. The authors employ a design‑for‑six‑sigma robust optimization of an outer‑rotor surface‑mounted permanent‑magnet motor for hybrid vehicles, using a ten‑dimensional Box‑Behnken response‑surface surrogate model coupled with particle‑swarm optimization to reduce computational cost. The surrogate‑based approach significantly improves optimization efficiency and effectiveness, with FEM validation confirming accuracy, and Monte Carlo simulations of mass‑production models demonstrate attainment of six‑sigma quality.
There is always uncertainty in industrial manufacturing. These uncertainties have an undesirable impact on the products if deterministic optimization approaches are employed. In order to have products as desired, uncertainties must be quantified in the design process. This paper presents a robust design optimization of an outer rotor surface mounted permanent magnet motor with particular application in the hybrid vehicle using the design for six-sigma methodology. Due to very long computational time of robust optimization, a ten high-dimensional surrogate model of the system using the Box-Behnken response surface methodology (RSM) is integrated with the particle swarm optimization (PSO). This causes a significant improvement in the effectiveness and efficiency of optimization. The results acquired from RSM are verified by their simulation using the finite-element method, and the accuracy of RSM is proved. Finally, the deterministic and robust optimized motors are simulated in mass production using Monte Carlo analysis, and six-sigma quality achievements are demonstrated.
32