Multi-level Design Optimization of an IPMSM Drive System Considering an Improved Loss Model

The worsening in oil resource crisis and climate change has contributed to the widespread investigations and applications of electric vehicles (EVs). Meanwhile, the performance of electrical drive systems, as the most important power unit, corresponds directly to the operational stability, reliability and economy of EVs. This paper proposes a system-level design optimization approach for the EV drive system with an interior permanent magnet synchronous motor (IPMSM), targeting improved system efficiency, reduced torque ripple and minimal costs. Specifically, an improved system-level loss model, considering both the motor and control system losses, is presented for objective efficiency calculation. Moreover, to handle the high dimensional and nonlinear problem, the multi-level optimization strategy is employed, in which four subspaces containing parameters in both motor and control levels are defined based on the sensitivity to the objectives. The approximate models and optimization algorithms are utilized to reduce the computation cost of finite element analysis in each level. Finally, the proposed approaches are investigated with a practical IPMSM drive system used in EVs. Results illustrate the effectiveness and show that the proposed multi-level design optimization method with the improved system-level loss model can obtain parametric optimums, while maintaining better steady-state and dynamic system performances.