针对刨煤机系统常规优化方法难以处理具有相互矛盾属性的系统可靠性多目标优化设计问题,提出一种基于遗传算法非概率分布信息的多目标模糊可靠性优化决策方法。该方法利用模糊集隶属函数的特性将各个单一目标函数模糊优化处理为能够反映各子目标相对重要程度,对不完全概率信息采用随机摄动法和Edgeworth级数方法转化为标准正态分布函数,将各个目标的协调满意度函数综合统一的单目标优化模型,再用常规优化方法即可求出多目标优化模型的满意解,最后给出刨煤机实例以验证该方法的有效性。

For a reliable multi-objective optimization design of the prow system, the conventional optimization method can not treat cases with mutually contradictory attributes. To overcome this difficulty, this paper proposes a multi-objective fuzzy reliability optimization decision method based on genetic algorithms and non-probability distribution information. The special features of the fuzzy set membership function are used to reflect the relatively important degree of the objective function and the probabilistic perturbation method and Edgeworth series technique are used to transform the incomplete probability information into the standardized normal distribution function, then a simple target model is established for each goal coordinated degree, and the conventional optimization method is used to obtain the multi-objective optimization satisfactory solution. Finally, a plow example is given to verify this method.