Aiming at the control circuit breaker temperature change in the manufacturing process,the process parameters were designed optimally. First,based on the uniformity of the better,more representative of the characteristics in the case that the number of uniform design and orthogonal design trials was similar,the test analysis was conducted. Then the regression model of temperature variations of circuit breaker was constructed by response surface methodology（ RSM）,and the residuals was utilized for interpolant fitting,which in turn to construct the response surface model based on radial basis function. Finally,an improved nonlinear genetic algorithm（ INGA） was proposed. In the selection of genetic algorithm,the elite strategy was firstly adopted to select the maximum fitness population,and then the roulette method was selected in the remaining population. The improved region was adjusted by the improved mutation operator,and using strong local search ability of the nonlinear programming algorithm,the accuracy of solution was upgraded. The results indicate that the improved nonlinear genetic algorithm has good global optimization capability,and it better than others.
Mechanical ＆ Electrical Engineering Magazine
radial basis function （RBF）
response surface methodology （RSM）
nonlinear-genetic algorithm （NGA）