PROCESS TRENDING WITH PIECEWISE-LINEAR SMOOTHING

Compression and trending are techniques widely used in the initial tre atment of raw process data. Two popular methods, Box Car with Backward Slope and Swinging Door, were reviewed recently by Kennedy. In spite of their popularity, neither of these methods are designed to cope wit h process variability and outliers. They also requre one or more param eters to be specified based on practical considerations. In this paper we propose a new method, Piecewise Linear Online Trending (PLOT), whi ch is statistically based and which performs significantly better. Unl ike the two existing methods, it adapts to process variability and noi sy data, recognizes and eliminates outliers, and it is robust even in the presence of outliers. It fits the data better for the same number of trends. The fidelity of its performance may be fine-tuned with a si ngle level of significant which may be set by the user without requiri ng any expertise in statistics. It may be used in an online or a batch mode, and interfaces easily with most existing packages.