Author:
Alekseeva A,Karpunina I,Klyachkin V
Abstract
Abstract
Under this consideration there is a process of potable water purification, where the chemical and physical characteristics (water turbidity and color, aluminum and chlorine content, etc) are monitored. To work with correlated characteristics it is necessary to apply to multivariate methods. The peculiarities of Hotelling charts plotting for individual observations are well known, but generalized variance charts plotting causes some problems connected with the assessment of covariance matrix for individual observations. Solving this problem one question arises: how often is it necessary to update the limit lines on the charts? We suggest applying to the test methods of the hypothesis on the equality of the mean.
Subject
General Physics and Astronomy
Reference11 articles.
1. A review of multivariate control charts;Lowry;Technometrics,1995
2. Hydraulic Unit Vibration Monitoring Using Machine Learning;Klyachkin,2020