针对产品质量特性稳健优化设计这一存在不确定因素的约束求解问题,提出一种基于量词约束满足(Quantified constraint satisfaction problem,QCSP)的稳健优化方法,将设计参数变差形式化描述为QCSP模型中的普遍性变量,在其作用下求得目标函数和约束函数的上下界,依据设计者偏好设定稳健性指标。根据模型特点,采取组合式算法进行求解,首先应用区间分析算法缩小设计变量搜索空间,获得满足普通优化设计约束的设计变量有效值域盒;然后由混合蛙跳算法在有效值域盒内搜索得到满足稳健性指标的帕累托优化解,并使用基于信息熵理论的方法选择出最优解。最后,应用该方法对透平压缩机扩压器进行质量特性稳健优化设计,证明了该方法在工程应用中的正确性和高效性。
Product quality characteristics robust optimization design is a constraint solving problem with uncertainties. A robust optimization method based on quantified constraint satisfaction problem (QCSP) is proposed, where, the design parameter variation is expressed as the universal variable. The upper and lower bounds of the objective functions and constraint functions are calculated with the effect of the universal variable, and the robustness indicators are set according to designer's preference. Considering the model features, a combination algorithm is applied to solve the QCSP model. Interval analysis algorithm is employed to narrow the search space of the design variables, thus valid domain box of design variables is obtained which satisfy the general optimization design constraints. The shuffled frog-leaping algorithm is used to search the Pareto optimal solutions which satisfy the robustness indices; in addition, the information entropy theory-based approach is applied to select the optimal solution. The efficiency and effectiveness of the proposed method are illustrated by the product quality characteristics robust optimization design of turbine compressor diffuser.