Integrated Washland Optimization Model for Flood Mitigation Using Multi-objective Genetic Algorithm


Abstract

Wash lands in river engineering have attracted attention as an alternative flood control facility and the promotion of biodiversity in a watershed. This article introduces an integrated washland optimization model that is developed to evaluate flood mitigation alternatives by installation of washlands in a large watershed. This model is aimed at encouraging collaboration between channels and basin storage for burden-sharing in the whole basin. The model's effectiveness in managing flood control is evaluated using a linkage between the optimization algorithm and Unsteady NETwork program (UNET) for hydraulic routing process and flood decrement computation in basin network. For optimal scheduling of the channel and washland capacity, two objective functions are defined: maximization of the freeboard in the channel, and minimization of the washland storage area. The multi-objective genetic algorithm (MOGA) based on a rank-based fitness assign method is used for optimization of these two objectives. The results indicate the optimal location of the washlands, and thus enable the precise evaluation of the freeboard increments by hydraulic routing process according to the pareto-optimal solutions in the basin-wide channel.