Control of Channel Bed Morphology in Large-Scale River Networks using a Genetic Algorithm

Alluvial channels undergo continuous morphological changes caused by relationships between entrained sediment, variable flows and movable boundaries. Excess changes that occur through sediment degradation and deposition, however, tend to threaten the stability of bridges, hydraulic control structures and underground utilities. These changes also reduce conveyance capacity of a channel and diminish reservoir benefits associated with hydropower generation, flood control, and water supply. This article outlines the development of an optimal control methodology for minimizing sediment aggradation and degradation, thus controlling channel bed morphology, in large-scale multi-reservoir river systems. The sedimentation control problem is solved by coupling the U.S. Army Corps of Engineer's HEC-6 sediment transport simulation model with an immune genetic algorithm. The simulation model is used to implicitly solve governing hydraulic and sediment constraints, while the genetic algorithm is used to solve the overall control problem. The method is demonstrated first through application to a hypothetical river network from the literature, for which a comparison between the genetic algorithm and alternative optimization technique is made. A second application to an existing hydraulic network illustrates the practical utility of the methodology as a decision-making tool for sedimentation control.