关键词: 不确定因素, 稳健性, 多目标, 支持向量机, 遗传算法

Abstract: To solve the performance reduction problem caused by uncertain factors existed in the process of support manufacturing and working,a robustness optimization method on machine tool support was proposed with the support static stiffness and weight as objectives.The objective function of robustness multi-objective optimization problem under the influence of uncertain factors was built based on the idea of interval method and robustness analysis,and a case study on a particular type of machine tool column was carried out with this method.The machine cutting point displacement under the loads action and the column weight were selected as the optimization objectives.The optimization variables were selected based on the sensitivity analysis of column feature sizes relative to the optimization objectives.With the column material density,material elasticity modulus and cutting loads as uncertain variables,the response models of optimization objectives related to optimization variables and uncertain variables were established.By adapting to IP-GA and NSGA-II algorithms,the column multi-objective optimization problem was solved.By comparing with the result of certainty optimization,the result obtained from the proposed method proved validity on support weight reduction and stiffness enhancement,especially on the stiffness robustness improvement (40% improvement compared with original column).

Key words: uncertain factors, robustness, multi-objective, support vector machine, genetic algorithms