Magnetic Field Analysis And Robust Design Of Segmented Halbach Cylindrical Permanent Magnet Synchronous Machines

The Halbach magnet array has potential features like providing sinusoidal fielddistribution, high flux density and self-shielding for its special magnetization, sincewhich more and more attraction has been paid to it in many electromagneticengineering domains such as electrical machines. The magnetic field is a medium tocovert mechanical energy to electrical energy or vice verse for an electrical machine.The understanding and prediction of the magnetic field distribution in electricalmachines are of great importance for the knowledge and comprehension of theirworking mechanism, the analysis and research of their electromagnetic performance,and the design and optimization of their structures. In this dissertation, the theory anddesign of segmented Halbach cylindrical permanent magnet synchronous machines(PMSMs) is mainly focused on, including the computation of the open-circuit andarmature-reaction magnetic field distribution, the prediction of the effect of statorslotting on the magnetic field distribution, and robust design.Based on the electromagnetic field theory, the analytical models for predictingthe magnetic field distribution in segmented Halbach magnetized machines aredeveloped. In order to ease the discussion and application, they are classified intointernal and external rotor machines having iron-cored or air-cored rotor. Predictedair-gap flux density distributions from the analytical models have been compared withfinite element (FE) calculations, and good agreement has been achieved. Further, therelationships between air-gap flux density and parameters, such as structureparameters and material properties of electrical machines, are investigated.Teeth and slots are usually distributed along stator bore in many PMSMs, whichdistorts the flux density distribution, and degrades the performance of PMSMs.Accurate analytical subdomain models for the open-circuit magnetic field in slotPMSMs are presented, accounting for the stator slotting effect. Comparison betweenanalytical and FE predictions illustrates that analytical models accurately reflect theinfluence of stator slotting effect, which makes the flux density waveform worse.Then, the back-electromotive force and cogging torque are obtained based on theanalytical models.The armature-reaction field has effects on the main field when a PMSM runs. Two analytical models for slotless and slot PMSMs are successively built forpredicting the armature-reaction field. The FE predictions confirm good accuracy ofthe two analytical models, and it is also shown that the slotting effect on thearmature-reaction field is significant. Based on the two models, the electromagnetictorque is calculated.The Taguchi method is a powerful tool for parameter design, which achievesrobustness through proper settings of certain parameters. Based on this method, theoptimal settings of structural parameters and material properties are identified toimprove the magnetic field distributions in segmented Halbach cylinders. Theexperimental results show that the effects of parameters and the optimal combinationsdepend on the problems, and it is necessary to weigh both advantages anddisadvantages and take all factors into consideration before making a decision.Particle swarm optimization (PSO) is a global optimization technique rooted inswarm intelligence, which has a few parameters and is easy to implement. TheTaguchi method is adopted to identify the settings of parameters for PSO since theyhave a large impact on optimization performance. A self-regulating mechanism and aself-evolving mechanism are introduced into PSO in succession to speed up theconvergence and improve the accuracy of solutions. Based on enhanced PSO, thecogging torque is minimized as the objective function, and the optimal settings ofparameters are identified for the robust design of a segmented Halbach cylindricalpermanent magnet synchronous machine.