Application of Numerical Algorithm Procedure to Classification of Expansive Soil

Due to the complexity of geological conditions and human factors of indoor physical and mechanical tests, evaluation of swelling potential of expansive soil shows significant characteristics of uncertainty, noise and multi-modal, Therefore its identification and classification is still a complex problem unsolved. This paper chooses liquid limit, plasticity index, clay content Less than 0.005 mm and free expansion rate as the evaluation indicators of expansive soil, after the determination of the value of each indicator, fuzzy C means algorithm model for level evaluation of expansive soil is established, then the author writes the evaluation process and uses this program on the cluster analysis of expansive soil. After compared with the indoor tests of expansive sample, the results indicate that fuzzy C means algorithm could achieve clustering of Nanyang expansive soil accurately and efficiently.