A New Coupled Optimization-Hydraulic Routing Model for Real-Time Operation of Regulated River Systems

This paper presents the River Simulation and Optimization Coupled Model (RSOCM) for the optimal operation of regulated river systems in real-time conditions. This model couples a highly robust and numerically efficient hydraulic routing approach with the well-known multi-objective Non-dominated Sorting Genetic Algorithm II (NSGA-II). The proposed model overcomes the lack of robustness of current models for unsteady flow routing. The lack of robustness of current unsteady flow models (e.g., numerical instability) constitutes a strong limitation for the development and implementation of real-time strategies in complex hydraulic systems that are intended to fulfill multiple objectives (e.g., minimization of flooding, optimal water allocation at specified diversion points). The real-time control of such complex systems may require thousands of computations of hydraulic routing for each operation interval. The optimization objectives supported by the RSOCM model include the optimal water allocation at specified diversion points and minimization of flooding. To demonstrate the proof of concept of the RSOCM model, it was applied to a hypothetical river system. As frame of comparison for the RSOCM model, the MODSIM-DSS model (Colorado State University, 1995) was used in this paper. The results suggest that the RSOCM model is a robust tool that can be potentially used for the optimal operation of multi-objective regulated river systems in real-time conditions.