The sinusoidal doubly salient electro-magnetic machine has the characteristics of simple structure

convenient magnetic field adjustment and small torque ripple. It can operate reliably under extreme conditions. A method for optimizing motor structure parameters was proposed based on parameter sensitivity analysis

which combines multiple regression equation and multi-objective genetic algorithm. The purpose is to solve the problem that the traditional optimization method of sinusoidal doubly salient electro-magnetic motor is difficult to establish the relationship between structural parameters and optimization objectives accurately and quickly due to the nonlinear characteristics. Taking the 12/10 sinusoidal doubly salient electro-magnetic machine with distributed magnetomotive forces as the research object

influence of structural parameter changes on the optimization objectives was explored by “sensitivity analysis method of simplified structural parameters”. Combined with the response surface method

the regression model between the motor optimization variables and the optimization objectives was obtained. Then

the optimal value is obtained by the single objective genetic algorithm and the normalized evaluation function. The finite element experimental results show that the back electromotive force of the optimized motor is increased by 14.94%

and cogging torque and total harmonic distortion are reduced by 89.25% and 17.5%. The experimental test further proves effectiveness of the design parameter optimization.