Unascertained Measure Model Applied on Identification and Classification of Expansive Soils

Based on the theory of unascertained, we propose a model to classify expansive soils, and to determine its expansive nature. The information entropy is adopted to determine the index weight, to establish functional relationship between the weight coefficient and the experimental values. The confidence level is also applied to identification and classification on expansive soils. Exemplification indicates that the evaluation result is consistent with actual instance by means of unascertained measure model. This method is convenient and practical compared with the fuzzy synthetical evaluation and gray correlation analysis.