Developing new tree expression programing and artificial bee colony technique for prediction and optimization of landslide movement

Abstract

The movement that occurs due to landslide is one of the most important issues in the field of geohazard. Determination of the movement of landslide is considered as a problematic task due to the fact that there are many effective parameters on movement of landslide that need to be investigated/observed carefully. In this study, various methods based on artificial intelligence were implemented and developed to evaluate and control this phenomenon. The gene expression programming (GEP) model is one of the newest models in artificial intelligence technique that can build the proper models for solving engineering issues based on the tree expression. Realistic data were used to design these models, where five model inputs including the groundwater surface, antecedent rainfall, infiltration coefficient, shear strength, and slope gradient of the area monitoring were considered as the input data. Many GEP models were constructed based on the most influential factors on GEP and according to two evaluation approaches, the best GEP model was selected. The obtained results of coefficient of determination (R2) for training and testing of GEP were 0.8623 and 0.8594, respectively, which indicate a high a capability of this technique in estimating real values of landslide movements. In optimization section of this study, artificial bee colony, as one of the powerful optimization algorithms was used to minimize risk induced by movement of landslide. According to the obtained results, no movement in landslide can be achieved if values of − 10.5, 400.1, 89.8, 59.65 and 24.95 were reported for groundwater surface, antecedent rainfall, infiltration coefficient, shear strength and slope gradient, respectively.