Abstract
In most multi-objective optimization problems of electrical machines, the weighted function method is used to convert them into single-objective optimization problems. This paper applies a kind of new multi-objective evolutionary algorithms (MOEAs), called adaptive multi-objective black hole (AMOBH) algorithms, to achieve effective multi-objective optimization of a tubular coreless linear permanent magnet synchronous motor (LPMSM). To reduce the computation cost of the MOEAs, a one-layer analytical model (AM) is presented for the tubular coreless LPMSM in this paper. The accuracy of the simplified one-layer AM is verified by comparisons with multi-layer AM and finite element analysis (FEA) under different structure parameters. It is found that the simplified AM has good accuracy and can decrease the computation cost significantly. AMOBH algorithm is subsequently introduced. The optimal Pareto front with regard to thrust, copper loss and permanent magnet volume are analyzed, and more diversified optimization results are provided. The final Pareto solution can be selected directly by practical physical values according to the application requirements. Finally, a prototype is fabricated for the selected design; its experimental results are provided and compared with those of the FEA results.
Funding Information
  • National Natural Science Foundation of China (61603358, 61703376)
  • Open Research Fund of Research Center for Advanced Control of Complex Systems
  • China University of Geosciences, Wuhan (AU2015CJ018)

This publication has 26 references indexed in Scilit: