A simulation-optimization method for ecohydraulic design of fish habitat in a canal

Abstract

A decision-support method for ecohydraulic design of a canal in rural engineering is developed using a GA (Genetic Algorithm)-based simulation-optimization technique. In order to demonstrate the ability of the method, a simple problem is presented aiming at searching for best configuration of blocks located for creating valuable fish habitat in a canal. The habitat condition is evaluated using a habitat suitability index (HSI) considering hydraulic factors such as water depth and velocity. The study area, a section of canal, is discretized into triangular finite elements with nodes, and the two-dimensional simulation of flow is conducted by the finite element method applied to the steady-state shallow water equations. The computed water depth and velocity at each node are then used to evaluate the weighted average HSI (AHSI) for the canal. In order to develop a robust method insensitive to uncertainty in these hydraulic parameters, a scenario-based stochastic optimization framework is adopted. The simulation part of the developed method comprises the numerical computation of canal flow at every scenario aiming at evaluation of the expected AHSI, whereas the optimization part includes rearrangement of blocks conducted by GA operators. The simulation and optimization process is iterated to maximize the expected AHSI in the canal section. The performance of the presented simulation-optimization method is demonstrated for a hypothetical ecohydraulic design of spawning area of Ayu (Plecoglossus altivelis) in a canal. The demonstrative example shows that the presented method can successfully specify the best block configuration in the canal section with the largest expected AHSI value.