ABSTRACT
The impermissible settlement is one of the most important reasons for the failures of earth and rockfill dams. An appropriate estimation of settlement after construction is required to evaluate the performance of the dam and to inform dam design engineers of any possible problem. This study was designed to apply artificial intelligent methods to predict settlement after constructing central core rockfill dams. Attempts were made in this research to prepare models for predicting settlement of these dams using the information of 50 different central core earth and rockfill dams all over the world using Genetic Programming (GP) methods. The results indicated that prediction of settlement based on the single parameter of dam height cannot be accurate and other parameters such as stiffness and shear strength properties of materials are effective in dam settlement. Based on the sensitivity analysis, parameters such as height of dam, modulus of elasticity and unit weight of core, and modulus of elasticity and internal friction angle of rockfill materials have the highest influence on the settlement of the dams and were considered as the input parameters. The results obtained from comparing the artificial intelligent method and empirical relationships showed that GP results were more appropriate tools to solve the problems with complex mechanisms and several effective factors, such as prediction of settlement of dams. The new developed models in this study are ready to be applied as a robust predictor tool for monitoring and safety evaluation of earth and rockfill dams in Europe.
