Profile monitoring assesses the stability over time of one or multiple functional quality characteristics to quickly detect special causes of variation that act on a process. In modern Industry 4.0 applications, a huge amount of data is acquired during manufacturing processes that are often contaminated with anomalous observations in the form of both casewise and cellwise outliers. Because anomalous observations can seriously affect the monitoring performance, profile monitoring methods that are able to successfully deal with outliers are needed. To this aim, we propose a new framework, referred to as robust multivariate functional control charts (RoMFCC), that is able to monitor multivariate functional data while being robust to both functional casewise and cellwise outliers. The RoMFCC relies on four main elements: (I) a univariate filter to identify functional cellwise outliers to be replaced by missing components; (II) a robust functional data imputation method of missing values; (III) a casewise robust dimensionality reduction; (IV) a monitoring strategy for the multivariate functional quality characteristic. An extensive Monte Carlo simulation study is performed to quantify the monitoring performance of the RoMFCC by comparing it with some competing methods already available in the literature. Finally, a motivating real-case study is presented where the proposed framework is used to monitor a resistance spot welding process in the automotive industry.
翻译:在现代工业4.0应用中,在制造过程中获取了大量数据,而制造过程中往往以不常见的外层和细胞外层两种形式受到异常观测的污染。由于异常的观测会严重影响监测性能,因此需要一种能够成功处理外部值的功能性监测方法。为此,我们提议一个新的框架,称为强有力的多变量功能性控制图表(ROMFCC),它能够监测多变量功能性数据,同时对功能性外端和细胞外端都保持稳健。ROMFCC依靠四个主要要素:(一) 一个单向过滤器,以识别功能性细胞外端,用缺失的部件取代;(二) 一种可靠的功能性数据估算方法,以缺失值进行成功处理。 (三) 以案例为主的稳健健的维度减少;(四) 多变量性功能性质量特征监测战略。开展了广泛的蒙特卡洛模拟研究,以量化RoMFCC监测性能。最后,通过将我们现有的一些现成本的研究,将一个现成本的研究加以比较。