Inversing mechanical parameters of concrete gravity dams using ant colony optimization

Abstract

Parameters inverse model of concrete gravity dams is usually formulated as an optimization problem with continuous variables, thus inverse procedures require optimization of an objective function. Ant colony optimization (ACO) is a novel technique proposed mainly for solving discrete optimization problems. In this paper, we attempt to present an adaptation and application of the ACO technique to the parameters inverse of concrete gravity dams. For this purpose, we firstly discretized the search space of mechanical parameters so that a parameters inverse problem was transformed into a combinatorial optimization problem. Secondly, according to the characteristics of the parameters inverse model, we adapted the ACO to the parameters inverse by redefining trail intensity and visibility. Lastly, we gave an example to test the adapted ant colony optimization (AACO) and the results show the AACO can optimize the inverse model validly and efficiently. © 2004 Springer-Verlag.