Fuzzy Expert System for Engineering Classification of Soils

Abstract

Soil classification is one of the most important stages in preliminary studies for analysis, design and applications in geotechnical engineering. The purpose of soil classification is to arrange various types of soils into groups according to their index and engineering properties and various other characteristics. Several existing agencies carry out the soil survey and classify the soil for different purposes. Fuzzy expert system is potent tool to mimic the human way of thinking and solve the problems dealing with uncertainty as they allow readable and interpretable rule bases. In this study, a fuzzy rule-based approach using fuzzy IF-THEN rules is used to quantify classification of soils in qualitative terms considering the index properties of soils. Triangular functions are used to fuzzify input properties of the soil to define the degree of membership. Fuzzy rules are generated to establish the soil types by aggregating fuzzified soil using the fuzzy operator. The experimental results are comparable with the outcomes of fuzzy system mode indicating the applicability of the developed fuzzy model for the engineering classification of soil.