基于进化计算的多目标鲁棒优化方法

工程应用中求解多目标优化问题时,所求的解既要具有较高的质量,又要满足指定的鲁棒性要求.对已有的多目标优化解的鲁棒性度量方法进行了分析,基于用户提出的严格鲁棒性要求,给出了一种严格鲁棒性度量方法并建立了求多目标鲁棒Pareto最优解的数学模型.模型归结为一个嵌套的双重优化过程,外层优化过程用于搜索高质量的解,内层优化过程用于测量候选解的鲁棒性度量指标.以进化计算作为搜索引擎,给出了实施模型的算法,仿真结果表明了方法是有效的.

Multi-objective robust optimization method based on evolutionary computation

The relatively high quality and appointed robustness of solutions are quite desirable in solving a multi-objective optimization problem of the engineering application.This paper proposes a strict robustness measure method and a mathematical model to obtain multi-objective robust Pareto optimal solutions based on user's strict robustness requirement and by analyzing the existing robustness measure methods of multi-objective optimization solutions.The model includes a dual nested optimization process,in which the external optimization process measures the performance of candidate solutions,while the inner optimization process measures the robustness index of candidate solutions.By using evolutionary computation as a search engine,an algorithm for implementing the model is proposed,and simulation results indicate that the proposed method is effective.