Modeling And Optimization Of Coaxial Magnetic Field Modulated Gearbox By Reluctance Network Method

With the merits of power transmission mode without mechanical contact,the magnetic gear has gradually regarded as a promising alternative of traditional mechanical gear.Therefore,a fast and accurate analysis tool is required to optimize the design of magnetic gears to commercialize this new technology.In this paper,a two-dimensional equivalent model of reluctance network for coaxial field-modulated magnetic gears is presented.The characteristic of the reluctance network model is that the magnetic field distribution in the reluctance network can be represented by an adjustable parameter system as well as the result is relatively accurate.In addition,without considering the magnetic saturation,the simulation time can be greatly declined without sacrificing the accuracy of torque prediction in the actual design.By consulting the literature,the design parameters are obtained.The structure and basic working principal of magnetic gears are mainly introduced.The analytical expressions of magnetic field of inner and outer air gap are proposed in detail and based on this theory the expressions of torque were deduced also.The reluctance network model is established by MATLAB programming.Firstly,the calculation methods of each element in the system network were presented.what is more,the mathematical representations of network model were emphatically described,then,the matrix model was solved and the magnetic field distribution was obtained.The accuracy of the magnetic reluctance network model was verified by comparing the predictions of the torque and air gap flux density component with the results of the non-linear finite element analysis and the simulation time is greatly reduced.Combined with the established reluctance network model,the effects of inner and outer air gap thickness,radial thickness of inner and outer rotor permanent magnets,parameters of magnetizing ring and arc angle of inner and outer rotor permanent magnets on the maximum static torque of magnetic gears were studied,and the design criteria of the above parameters were given.At the end of this paper,a new optimization method of magnetic gears with fast convergence and accurate calculation is attained by combining the equivalent reluctance network model with the multi-verse algorithm.This method can optimize many parameters globally at the same time,it ban be applied to the preliminary selection of parameters in engineering practice.