LAUSR

To improve the precision of the magnetostatic finite element model for the long-stroke moving-iron proportional electromagnetic actuator, on the basis of three characteristic parameters, including the initial relative permeability of materials μr, saturation magnetic polarization JS, and knee point adjustment coefficient of magnetizing curve a, this article puts forward an efficient demarcation method by virtually remodeling the magnetizing curve of soft magnetic material for the actuator and combining surrogate model and optimization technique. Then, the demarcation mathematical models based on surrogate model are built while taking μr, JS, and a of the soft magnetic material 1 for the main magnetic loop of the actuator as calibration parameters, and regarding the relative root-mean-square error and relative maximum absolute error of electromagnetic force under the typical operating conditions of the actuator as targets. Finally, the demarcation mathematical models are solved by downhill simplex method and the final calibration parameters are confirmed. The results indicate that the relative root-mean-square error, relative maximum absolute error, and relative average absolute error of finite element model after demarcation are 2.7, 4.3, and 2.3%, respectively, decreasing of 57.1, 58.7, and 39.5% with respect to the initial finite element model of the actuator.