Optimization of an induction motor for loss reduction considering manufacturing tolerances

As the core component of electric vehicles (EVs), the performance of electric motors directly affects the economy and safety of EVs. Manufacturing tolerances will cause the actual performance parameters of the motor to deviate from the expected value, resulting in a reduction in motor service life. This paper presents a reliability-based design optimization (RBDO) method for three-phase squirrel-cage induction motor (SCIM) to manage the manufacturing tolerances and reduce its influence on the motor total loss and starting performances. First, the performance analysis based on equivalent circuit method (ECM) that focuses on the influences of rotor dimensions on the loss and starting performances is presented. In addition, surrogate models for the total loss, starting torque, and starting current are constructed based on candidate designs and corresponding responses obtained via electromagnetic finite element analysis (FEA), and the model accuracy is assessed by evaluation indexes. Then, the adaptive-weighted response surface method (AWRSM) is applied to adaptively approximate the limit state function, and Monte Carlo Simulation method is adopted for reliability analysis. Finally, a RBDO of a SCIM for loss reduction considering manufacturing tolerances is established. The final optimization result suggests that the total loss is reduced by 8.1 W compared to the initial value. Compared to the deterministic optimization results, the reliability of starting torque and starting current is improved by 51.93% and 47.76%, respectively.