Robust multiobjective and multidisciplinary design optimization of electrical drive systems

Design and optimization of electrical drive systems often involve simultaneous consideration of multiple objectives that usually contradict to each other and multiple disciplines that normally coupled to each other. This paper aims to present efficient system-level multiobjective optimization methods for the multidisciplinary design optimization of electrical drive systems. From the perspective of quality control, deterministic and robust approaches will be investigated for the development of the optimization models for the proposed methods. Meanwhile, two approximation methods, Kriging model and Taylor expansion are employed to decrease the computation/simulation cost. To illustrate the advantages of the proposed methods, a drive system with a permanent magnet synchronous motor driven by a field oriented control system is investigated. Deterministic and robust Pareto optimal solutions are presented and compared in terms of several steady-state and dynamic performances (like average torque and speed overshoot) of the drive system. The robust multiobjective optimization method can produce optimal Pareto solutions with high manufacturing quality for the drive system.