论文部分内容阅读
针对质量特性为轮廓(Profile)的输出响应的优化问题展开研究,提出一种基于主成分分析的双响应曲面法和满意度函数相结合的函数响应优化方法。将Profile的每个观测点看成一个独立响应,将Profile问题转化为多响应问题。求得多个观测点的均值和方差的满意度函数值,通过主成分分析法,将多个观测点的均值和方差的满意度函数值转化为主成分综合得分,并将这两者的加权和作为最终的优化指标。本文所提方法可以有效解决观测点之间存在的相关性的问题,并且优化过程同时考虑到每个观测点响应的均值和方差影响。实例证明,该方法简单易行,优化结果满意。
In order to solve the problem of optimization of the output response of profiles, a method based on principal component analysis (PCA) is proposed to optimize the function response based on the dual response surface method and the satisfaction function. Think of each observation point of the Profile as an independent response, transforming Profile questions into multiple response questions. The mean value and variance satisfaction function values of multiple observation points were obtained. Principal component analysis (PCA) method was used to convert the satisfaction and satisfaction of multiple observation points into the main component comprehensive score, and the weights of the two And as a final optimization indicator. The proposed method can effectively solve the problem of the correlation between observation points, and the optimization process takes into account the mean and variance of the response of each observation point. The example proves that the method is simple and easy to be implemented and the optimization result is satisfactory.