Parameter setting and optimization are key steps to improving the transmission performance of eccentric permanent magnetgear. Studying this step makes it possible to analyze the transmission relationship between the design parameters, and improve both...
Parameter setting and optimization are key steps to improving the transmission performance of eccentric permanent magnetgear. Studying this step makes it possible to analyze the transmission relationship between the design parameters, and improve both the operation stability and the carrying capacity of eccentric permanent magnet gear. Taking multi-shaft ring-plate magnet gear (MRMG) as the research object, this paper analyzes the influence of actual structural parameters on torque performance through the comprehensive sensitivity method, and selects the parameters with the greatest influence as the main optimization parameters. Then using the BSO algorithm and the BP neural network, a prediction model of MRMG electromagnetic torque and torque density is established, the relationship between each of the parameters is analyzed, and the optimization parameter points are determined. By comparison with finite element simulations and experiments, the built model and its measured results and FEM analysis are shown to have good accuracy (error < 6%). Moreover, the model can quickly and accurately obtain optimization parameters, and greatly improve the transmission performance of the structure (the maximum torque increases by 15%, and the torque density increases by 36%).