基于反馈控制原理可靠性多目标优化设计的满意解

通过对多目标优化问题不确定性因素的分析，将有效解问题扩展到模糊多目标优化模型软计算满意解求解问题．利用模糊集理论中隶属函数，将各子目标模糊化后处理为满意度函数，进而在分析控制系统中的闭环控制原理和多目标满意解优化软计算的动态过程的相似性后，提出了建立基于闭环反馈控制原理，由协调校正权函数、相对优属度函数和重要性加权指数构成的多目标满意解总评价函数方法．最后，给出机械零件可靠性多目标满意解优化实例进行验证．

The Satisfaction Solution of Multi-objective Optimization of Reliability Based on Control Principle of Feedback

Through analyzing the uncertainty factors of multl-objective optimization problems, it shows that Pareto solution should be generalized to fuzzy multi-objective optimization model, and the model is used to find the satisfactory solution through soft computing. According to the property of membership function in fuzzy sets theory, each sub-objective can be dealt with satisfactory function. After analyzing the similarity between closed loop control theory in control system and the dynamic progress in soft computing of multi-objective satisfactory solution optimization, the paper developed a general judgment function method of multi-objective satisfaction solution. The method is based on closed loop feedback control theory and is dealt with correcting weight function, relative optimum membership function and weight index of importance. At last, an illustrative example of multi-objective optimization of mechanical component reliability was given.