Research On The Method Of Electromagnetic Noise Suppression Based On The Optimization Of The Structure Parameters Of The Electric Vehicle Driving Motor

With the rapid popularization and technological progress of electric vehicles,the electromagnetic noise problem of driving motor is increasingly affected by the development trend of motor technology,such as lightweight,high speed and high torque density.Compared with motor hydrodynamic noise and mechanical motion noise,electromagnetic vibration noise has a higher amplitude,which is more likely to cause discomfort and more difficult to suppress.Structural parameter optimization is a necessary part of electromagnetic noise suppression.The key lies in how to obtain the structural parameters to be optimized through the sensitivity analysis of various,miscellaneous and coupled motor structural parameters for the electromagnetic noise and output performance and how to optimize these parameters in an orderly,effective and efficient manner.Therefore,a sensitivity analysis method based on the function classification of structural parameters and the metamodel of optimal prognosis and a hierarchical and progressive optimization method of parameters to be optimized were proposed:Firstly,the radial electromagnetic force density was analyzed based on Maxwell stress tensor method to clarify its composition and causes.At the same time,the key components of the radial electromagnetic force density were determined based on the spatial and temporal characteristics of the radial electromagnetic force density with low spatial order and high amplitude,which causes strong electromagnetic noise.Based on air-gap permeability and magnetic potential analysis,the mechanism of restraining torque fluctuation caused by tangential air-gap excitation was analyzed.A parameterized finite element electromagnetic calculation model for the whole structure of the driving motor was established,and the structure of the motor was represented by 46 structural parameters.Based on air gap electromagnetic excitation and motor output performance,the calculation content of finite element model was determined.The calculation accuracy of the electromagnetic performance of the model was verified by the motor bench experiment,and the working points and causes corresponding to the strong electromagnetic noise were identified.A sensitivity analysis method based on structural parameter action classification and the metamodel of optimal prognosis was proposed.Firstly,the parameters were classified based on their function for the motor structure,and the relation between the various analysis results was established by the referring parameters.Secondly,the relationship model between the structural parameters and the objective function was fitted based on the metamodel of optimal prognosis,and the sensitivity level of the structural parameters was evaluated quantitatively based on the contribution of the parameters to the fitting accuracy,so as to realize the screening and grading of the parameters.In order to realize the orderly,effective and efficient optimization of the parameters to be optimized after the sensitivity screening and classification,a hierarchical and progressive optimization method was proposed.According to the sensitivity analysis results,the structural parameters were optimized according to the order of strong,medium and weak sensitivity levels.Each level of optimization took the superior optimization results as input,so as to realize grading and progressive.Meanwhile,the optimization method was adjusted according to the number of parameters to be optimized and the sensitivity level.The optimization results showed that the electromagnetic excitation of the air gap was suppressed and the sound pressure level of electromagnetic noise was reduced without reducing the output performance of the motor.