Comprehensive multi‐objective scalarisation optimisation of a permanent magnet machine with correlation parameters stratified method. Issue 1 (1st January 2017)

• Record Type: Journal Article
• Authors: Yin, Jianing; Zhu, Xiaoyong; Quan, Li; Zhang, Chao; Xiang, Zixuan
• Is Part Of: IET electric power applications. Volume 11:Issue 1(2017)
• Page Start: 72
• Page End: 79
• Publication Date: 2017-01-01
• Subjects: permanent magnet machines -- Pareto optimisation; multiobjective scalarisation optimisation -- permanent magnet machine -- correlation parameters stratified method -- high‐quality motor driving system -- high‐dimensional optimisation problems -- Pareto optimal points -- weight factors; Electric power -- Periodicals; Electric power systems -- Periodicals; 621.305
• DOI: 10.1049/iet-epa.2016.0300
• Languages: English
• ISSNs: 1751-8660

Abstract

Abstract : With the increasingly high requirements in high‐quality motor driving system, the multi‐objective optimisation of permanent magnet (PM) machines has become a hot issue in recent years. However, such multi‐objective optimisation is often concerned with high‐dimensional optimisation problems and the obtained Pareto optimal points are non‐single, hence it is relatively complex and difficult to obtain a definite result in the design of PM machine. In this study, a comprehensive multi‐objective scalarisation optimisation approach with the correlation parameters stratified method is proposed, where not only the impacts of each design parameter on objectives are investigated, but also the interactions among the design parameters themselves are considered and analysed in detail. In addition, by introducing the weight factors of the desired objectives, the multi‐objective optimisation is transformed into a simplified scalar optimisation problem. Then, an interior PM machine with L‐shaped PMs is investigated by using the proposed method. Finally, based on the optimised result, a prototype machine is built and tested. Both simulation and experiment results verify the validity of the proposed method.