| As price rising and supply shrinking of the rare-earth material market,the permanentmagnet-assisted synchronous reluctance machines(PMaSynRMs)have attracted significant attention and possess broad application prospect,due to their low cost,small size,high torque density,high efficiency,wide speed range,no rare-earth material consumption,and high reliability in the harsh thermal environment.However,the inherent demerits of the PMaSynRMs(e.g.,low power factor,high torque ripple and vibration noise,complex rotor structure,difficulties in motor parameterizing,and time-consuming design process)severely restricts the further promotion and application in the industrial fields.Hence,this thesis focuses on the comprehensive design and optimization of the PMaSynRMs,including the basic theory,rotor flux barriers,fast motor parameterization method,and automatic optimization method of the PMaSynRMsFirstly,this thesis thoroughly analyzes the structural characteristics,mathematical model,and operating principles of the PMaSynRMs,to lay the theoretical foundation below.Then,the design of the rotor flux barriers is studied.And the relevant design basis is established by analyzing the magnetic field distribution of the solid rotor and the theory of hydrodynamics,which is summarized as follows: To maximize the rotor anisotropy and increase the torque density of the PMaSynRMs,the rotor flux barrier should be reasonably designed to make its edges coinciding with or paralleling to the rotor q-axis magnetic field lines,so that the flux barrier can block the rotor d-axis magnetic flux to the hilt.Finally,the Airfoil Potential Function is introduced to fit the distribution of the q-axis magnetic field lines of the solid rotor.Based on this,the field-line shaped(FLS)barrier with the fullest potential of maximum rotor anisotropy is proposed.For the difficulties of comprehensive performance optimization of the PMaSynRMs,this thesis proposes a general method for fast rotor parameterization.And take C-Shaped,U-shaped,and FLS barriers as the examples,the implementation steps of the method are clarified in detail.The difference between the three parametric models is also discussed,and the parametric model of the FLS barrier rotor is the most concise and has the least degree of freedom.Afterward,this thesis intensely studied the principles,merits,and implementation methods of the multiobjective genetic algorithm(MOGA).The relevant results show that MOGA is vastly suitable for solving the optimization problem of the PMaSynRMs,due to its excellent ability of global search,strong stability,fast converging rate,and low calculating complexity.Finally,based on the above research results,an automatic design and optimization platform for PMaSynRMs is designed by combining with FEMM,which is an open-source Matlab toolbox for finite element analysis(FEA).The platform is integrated with many rotor and stator geometries so that it has great versatility in the design of PMaSynRMs;Besides,the platform is integrated with various functional program scripts,so that it can perform the electromagnetic performance analysis and the results extraction automatically by reading the machine design parameters and instructions entered by the users;Furthermore,the platform is integrated with the m-language-based MOGA script that linked naturally with the FEMM toolbox,so that it can achieve the comprehensive performance optimization of the PMaSynRMs efficiently and fastly.Based on the proposed platform,the design optimization of a 22 k W PMaSynRM with FLS barriers for the air compressor application is performed.Firstly,the considerations of slot-pole combination,barrier shape determination,and material selection are discussed according to the design specifications of the target machine.Then,based on the rotor “virtual slot” principle,several candidate models with different slot-pole combinations and rotor barriers are designed and analyzed.The research results show that the 36-slot-4-pole PMaSynRM with 4-layer FLS barriers has the highest rotor design flexibility,saliency ratio,torque density,and power factor under the geometry constraints of the target machine.Hence,it is determined as the initial design model.After that,based on the optimization module of the platform,the multi-objective optimization of the initial machine model is successfully achieved when the optimization objectives are selected as average torque and torque ripple.Notably,three optimization tricks are adopted to boost the optimization process,which are summarized as follows: limiting the threshold of optimization objectives,designating specific individuals for the initial population,and setting the current phase angle as one optimization parameter.Finally,the structure and electromagnetic performance characteristics of the final model and the initial model are compared and analyzed,and the structural reliability of the motor rotor is also demonstrated. |