Year
1979
Abstract
Facts and falacies of process tank calibration concern minimization of modeling error associated with a tank and the use of redundancy to determine the state of stability of the measurement system. Redundant information on the performance of the measurement process provides a means to eliminate outliers from the database. A technique is presented which minimizes the modeling error and, therefore, systmatic components of the variance and allows estimates of the within-sequence and betweensequence variabilities.