Study of the PSA-ANFIS-based inverse design approach on geotechnical engineering

Abstract

The design work is important to normal employment and safety and stabilization of a project in geotechnical engineering. The system researched in geotechnical engineering is altitudinal complex and indeterminate, so traditional design methods are difficult to deal with an actual project ideally. According to the characteristics of inverse design method in geotechnical engineering, the problems on setting the training parameters and model of adaptive neuro-fuzzy inference system (ANFIS) was studied by using orthogonal test design method and pattern search algorithm (PSA), and then the approach based on PSA-ANFIS was put forward and established. Furthermore, data pairs for training, data pairs for checking and data pairs for prediction were built by using a multi-peaks function to analysis the training process and the fitting and predicting capability of the PSA-ANFIS. The results show that the PSA-ANFIS is well in parameters adjustment and model establishment, and that is an excellent approach to deal with the nonlinear and complex fitting problem of inverse design method in geotechnical engineering. Finally, the PSA-ANFIS-based inverse design approach is applied to inverse design the angles and heights of some slopes. The result shows that the new approach has ability for handling the nonlinear problem, and is suitable for dealing with the altitudinal complex and indeterminate problem in geotechnical engineering, and is an excellent inverse design approach.