Multi-objective optimization design of novel tubular switched reluctance motor

PurposeThe purpose of this paper is to achieve the multi-objective optimization design of novel tubular switched reluctance motor (TSRM). Design/methodology/approachFirst, the structure and initial dimensions of TSRM are obtained based on design criteria and requirements. Second, the sensitivity analysis rules, process and results of TSRM are performed. Third, three optimization objectives are determined by the average electromagnetic force, smoothing coefficient and copper loss ratio. The analytic hierarchy process-entropy method-a technique for order preference by similarity to an ideal solution-grey relation analysis comprehensive evaluation algorithm is used to optimize TSRM. Finally, a prototype is manufactured, a hardware platform is built and static and dynamic experimental validations are carried out. FindingsThe sensitivity analysis reveals that parameters significantly impact the performance of TSRM. The results of multi-objective optimization show that the average electromagnetic force and smoothing coefficient after optimization are better than before, and the copper loss ratio reduces slightly. The experimental and simulated results of TSRM are consistent, which verifies the accuracy of TSRM. Research limitations/implicationsIn this paper, only three optimization objectives are selected in the multi-objective optimization process. To improve the performance of TSRM, the heating characteristics, such as iron loss, can be considered as the optimization objective for a more comprehensive analysis of TSRM performance. Originality/valueA novel motor structure is designed, combining the advantages of the TSRM and the linear motor. The established sensitivity analysis rules are scientific and suitable for the effects of various parameters on motor performance. The proposed multi-objective optimization algorithm is a comprehensive evaluation algorithm. It considers subjective weight and objective weight and fully uses the original data and the relational degree between the optimization objectives.