基于可能性准则的一元累积和控制图(英文)

An univariate CUSUM control chart based on possibility rule

基于过程检验统计量模糊化后的隶属函数确定的可能性分布,提出了一种对质量模糊特性进行累积和控制的方法,其中应用了一类新的距离概念、新的方差和样本方差.将LR-型质量数据看作中心值变量服从正态,二扩展变量服从均匀分布的模糊随机变量,进而对过程均样本作近似正态处理,根据Kruse-Meyer对模糊数据假设检验的方法,给出了独立观测下一元累积和控制的可能性准则.

We propose a cumulative sum control method based on possibility distribution determined by the membership function of fuzzified test statistics for quality control chart centering on fuzzy items of products quality, where a sort of novel distance, variance and sample variance is applied, and the LR-quality data is viewed as a fuzzy random variable with normally distributed central variable and two uniformly distributed spread variables, so that the process mean sample can be treated as approximately normal. According to hypothesis testing methods for fuzzy data proposed by Kruse-Meyer, a possibiliy rule for univariate Cumulative Sum （CUSUM） control chart under independent observations is presented.