Application of artificial intelligence-based single and hybrid models in predicting seepage and pore water pressure of dams: A state-of-the-art review

Abstract

Failure of earth dams is one of the major challenges of civil engineering, one of the main causes of which is uncontrolled seepage from the core and foundation of the dam. The use of numerical methods, analytical methods, and other modeling methods in solving the problem of dam seepage and pore water pressure is common, but in recent years, the use of artificial intelligence (AI) models and hybrid methods have specifically identified for this purpose. The results of a review study of artificial intelligence models in predicting leakage and pore water pressure of dams show that machine learning (37.53%), neural network (27.63%), and hybrid models (21.05%) are more popular than other techniques. Single models artificial neural networks (ANN), support vector regression (SVR), random forest (RF), and feed forward neural network (FF-NN) have been used more than other models. Also, 81.25% of the hybrid models have used neural network models. Also, 31.25% of the models have used the genetic algorithm (GA) in their hybrid model. Accordingly, 46 research papers from 2005 to 2022 were reviewed. This review was conducted employing preferred reporting items for systematic reviews and meta-analyses (PRISMA) method. The present review article provides comprehensive research on the application of intelligent models to model the seepage and pore water pressure of dams and provides in-depth insights into the use and validity of different modeling methods for dam seepage.