An Adaptive Network Fuzzy Inference System Approach for Site investigation

Abstract

Abstract Site investigation has to be effective and must be carried out in a systematic way. The purpose of this article is to evaluate the number of investigation points, field tests, and laboratory tests for a description of a building site. Such an assessment depends on many parameters based on experiences which cannot be physically evaluated. The guidance on spacing is available from many sources, and such guidance provides a starting point for the extent of investigation. The recommendations were examined and used for building of the model to predict the optimal number of investigation points. Several parameters with the biggest influence on the number of investigation points were considered. The influence of each parameter was determined on the basis of recommendations and engineering judgment. Increments of the minimum number of investigation points for a different building site conditions were used to construct the model with adaptive network fuzzy inference system (ANFIS). The formed ANFIS-SI model was applied on reference cases. There is a good agreement between the model and the reference cases. Additionally, the recommendations for the type and frequency of tests in each stratum are provided to optimize the soil investigation. The ANFIS-SI model, with integrated recommendations, can be used as a systematic decision support tool for engineers to evaluate the number of investigation points for a description of the building site.