ESTIMATION OF THE LONGITUDINAL DISPERSION COEFFICIENT FOR RIVER NETWORKS USING A DIFFERENTIAL EVOLUTION ALGORITHM

Parameter estimation for river network systems is an important issue in environmental science and has attracted increasing interest from various research fields, it could be essentially formulated as a multidimensional optimization problem. As a novel evolutionary computation technique, differential evolution (DE) algorithm has attracted much attention and wide applications owing to its simple concept, easy implementation and quick convergence. In this paper, a new parameter identification model based on a DE algorithm coupled with a water quality model was constructed for the determination of longitudinal dispersion coefficients for river networks. It concluded that a DE algorithm has better global convergence than a standard genetic algorithm in the solution of De Jong function F2. The method was validated by a river network numerical test which was composed of nine channels. The computational results indicated that the model could give good identification precision results.