二元非对称质量损失的三角分布过程均值优化

对于非对称质量损失，过程均值并不是越接近目标值越好，需要对过程均值进行优化．针对三角分布这种常见的非正态分布，研究了非对称质量损失时其过程均值的优化问题．在总结田口提出的二元非对称质量损失函数的基础上，分析了质量特性为三角分布时过程均值的变化对质量损失的影响，并指出过程均值变化的4种情况；给出了计算非对称质量损失的数学模型，并得到三角分布时最佳过程均值的计算方法．最后，用实例验证了此过程均值优化模型的有效性．

Optimizing the Process Mean of Triangular Distribution under Quadratic Asymmetric Quality Loss

It is necessary to optimize the process mean under asymmetric quality loss as the process mean is not the closer to the target value the better. The issue of determining the optimum process mean with asymmetric quality loss was studied for common non-normal triangular distributions. The impact of process mean changes of the triangular distribution on the quality loss was thoroughly analyzed based on the summary of Taguchi＇s quadratic asymmetric quality loss function, and four situa- tions for changing the process mean were derived. Mathematical models for calculating the triangular distribution under the four different conditions were introduced. Calculating methods for the four optimizing models of the triangular distribution were studied. The validity of the process mean optimizing models was verified with examples.