Statistical control of multidimensional scattering from individual observations

Abstract

The features of the multi-parameter process control based on individual observations, when sampling during monitoring is impossible due to the complexity of measurements or is not economically feasible, are considered. Monitoring of multidimensional scattering usually by using the generalized variance algorithm is carried out, however, when controlling for individual observations, a number of problems arise. An approach to estimating the covariance matrix to solve this problem is proposed. Keywords multivariate statistical control, generalized variance, sliding range