Design Optimization Research Of The Double Salient Permanent Motor Based On Sequential Unconstrained Minimization Method

Double Salient Permanent Magnet (DSPM) Machine is a new type electrical machine which has been thoroughly studied in recent years. However the design optimization, as an important procedure in the development of electrical machine, has seldom been researched. In this thesis, the research will be developed into it.Design optimization of electrical machines is a nonlinear constrained and mixed-discrete multiple variables optimization problem. Based on the investigation to the theory of motor optimization design, the SUMT (Sequential Unconstrained Minimization Method), which is widely used in motor optimization, and combine Pattern Search and Improved Powell Method are selected to optimize the design of DSPM.In this thesis, firstly the fundamental principle of DSPM is introduced. Then the mathematical model of DSPM optimization design is established. After that, SUMT (Based on Pattern Search) and SUMT (Based on Improved Powell) are fully discussed. For each algorithm, we study its principle and give the detailed procedure for DSPM optimization. Finally, a software for DSPM optimization design has been developed. These two algorithms are used to optimize selected DSPM machine and the results are accumulated and analyzed. The results show that the performance of both algorithms is good and has some practical value.