| 摘要 |
<p>A method of handling, as bulk material, a substance which is one member of a class of multi-component substances includes subjecting successive test samples of the substance to a series of measurements which provide multiple data point which are influenced by at least one selected parameter and from which are determined the factor loadings for the respective test sample of each of multiple pre-determined abstract factors. Pre-determined regression coefficients are utilized to calculate from the factor loadings a predicted value for the selected parameter for the respective test sample. The predicted value is compared with one or more related set points, and the constitution and/or disposition of the substance is controlled in response to the outcome of the successive comparisons. The regression coefficients have been determined by correlating knowns values of the selected parameter for standard characterized samples of each of multiple substances of the class, with factor loadings derived by factor analysis of multiple data points which were obtained by subjecting the standard characterized samples to the series of measurements. </p> |
