Clustering Optimization of IPMSM for Electric Vehicles: Considering Inverter Control Strategy

Authors:
Open Access
Publisher website
Google Scholar
Abstract
The actual performance of driving motors in the electric vehicle (EV) powertrain depends not only on the electromagnetic design of the motor itself but also on the driving condition of the vehicle. The traditional motor optimization method at the rated point is difficult to deal with because of the mismatch between its high-efficiency area and the actual operation area. This paper systematically proposes an optimal design method for driving motors for EVs, considering the driving conditions and control strategy to improve motor efficiency and passengers’ riding comfort. It uses cluster analysis to identify representative points and related energy weights to consider motors’ comprehensive performance in different driving cycles. Three typical operation conditions are selected to implement the proposed optimization process. In the design process, by using the sensitivity analysis method, the significance of the structural parameters is effectively evaluated. Moreover, the semianalytical efficiency model and torque model of permanent magnet driving motors based on finite element analysis results are deduced to consider the influence of magnetic saturation, space harmonics, and cross-coupling between d-axis and q-axis magnetic fields. Based on the driving system demands of an A0 class pure EV, the whole optimization design is divided into four steps and three scales, including the motor scale, control scale, and system scale. By using the multi-objective optimization method, Pareto optimality of motor efficiency and torque ripple is achieved under the city driving cycle and highway driving cycle. Compared to the optimization only at the rated condition, the proportion of motor sweet region increased about 1.25 times and 3.5 times by the proposed system-scale optimization under two driving cycles, respectively. Finally, the effectiveness of the proposed optimization method is verified by the prototype experiments.
Funding Information
  • National Natural Science Foundation of China (52277065, 2020YFA0710500)
  • National Key Research and Development Program of China (52277065, 2020YFA0710500)

This publication has 37 references indexed in Scilit: