Siyue Wu / College of electrical engineering; Zhejiang University
This paper presents a hybrid optimization method based on the multi-Support Vector Machine (multi-SVM) and Kriging method for linear induction motors (LIMs). With this novel method, the optimal schemes can satisfy various performance specifications of LIMs after multi-objective optimization process. The raw data including structural parameters and performance indexes are obtained from finite element analysis (FEA). Utilizing multi-SVM, a global analysis model is established to observe the optimization trend under entire reference interval. With the preliminary assessment and range selection, a high-precision interval model based on Kriging method enhances the efficiency and accuracy of analysis. The adjustable objective function ultimately determines the design scheme and realizes the optimization objectives of the thrust force, thrust ripple, vertical force, current, power factor, copper loss and weight simultaneously.
