河网与闸泵站工程耦合水动力模型HydroNet:模型构建与验证OA

Traditional hydraulic models developed for complex river network systems often lacked detailed consideration of controlling hydraulic structures,leading to limitations in accurately representing flow regulation mechanisms.However,driven by the increasing demand for smart water management,the refinement of hydraulic component modeling within these systems became essential to support real-time control and decision-making.To address this challenge,a coupled hydrodynamic simulation model,HydroNet was independently developed to integrate river network hydraulics with detailed hydraulic structure dynamics.This paper introduces the model's theoretical framework,numerical methods,and validation cases. In the HydroNet model,hydraulic components including channels,sluice gates,pumps,turbines,and inverted siphons were systematically integrated,with each component connected to an upstream and a downstream hydraulic node.These nodes,in turn,could link to multiple components,enabling the construction of complex river network topologies with arbitrary configurations.To solve the water level at each node,a mass balance equation was formulated by accounting for the inflow and outflow contributions from all connected components.The formulation of the nodal water level equations required explicit relationships between water level and discharge at both the upstream and downstream ends of each component.For channels,this relationship was derived from the Saint-Venant equations,discretized using the Preissmann finite difference scheme to ensure numerical stability and accuracy.For control structures such as sluice gates,pumps,turbines,and inverted siphons,their empirical discharge equations linearized via first-order Taylor expansion were incorporated into the nodal system.Furthermore,the flow diversion outlets was treated as internal boundary conditions,where diverted flows were directly coupled to nodes as specified outflows in the mass balance equations.As for the external boundary conditions,which is prescribed as time-varying water levels or discharge time series,were imposed at the periphery of the river network system to close the mathematical framework. The coupling of these hydraulic relationships allowed the HydroNet model to assemble a unified global sparse linear matrix representing the entire river network.The sparse matrix could be solved by highly efficient direct matrix solvers,where the water level of all nodes at the next time step can be solved simultaneously.On this basis,taking the node water level as the boundary condition of the channel,the water level and flow rate of all sections in the channel at the next time step can be directly calculated.Taking the node water level as the upstream and downstream conditions of the gate,pump,turbine and inverted siphon,the flow rate of these controlling hydraulic structures at the next time step can also be directly calculated,and thus the time step of the entire river network hydraulic calculation is completed.This approach ensured proper handling of bidirectional flows,backwater effects,and abrupt transitions induced by hydraulic structures. A key feature of HydroNet was its ability to construct and solve unified nodal equation systems for river networks with arbitrary topological configurations.The model supported both series and parallel arrangements of hydraulic structures,enabling flexible simulation of real-world water distribution systems.Node equations were derived based on mass conservation.Additionally,the model incorporated advanced numerical treatments for challenging flow conditions,such as wetting-drying transitions and free surface-pressurized flow alternations,ensuring robustness in unsteady flow simulations. To evaluate the model's performance,three validation cases were conducted.The first two benchmark tests focused on the model's capability to simulate dry-bed flooding and complex branching river networks,demonstrating its stability and accuracy in handling abrupt flow changes.The third case study applied HydroNet to a 146 km segment of the South-to-North Water Transfers Project,which included multiple hydraulic structures such as sluices,pumps,and bifurcation nodes.Simulation results closely matched field measurements,confirming the model's reliability in large-scale,structure-intensive scenarios. The proposed HydroNet framework significantly expanded the applicability of conventional river network models by introducing high-resolution hydraulic structure coupling.Its successful implementation demonstrated that integrated hydrodynamic modeling could effectively support smart water management applications requiring real-time system regulation.Future work will focus on enhancing computational efficiency and integrating remote sensing data for adaptive model calibration.The research significance of this model lies in enhancing the practicality of river network hydrodynamic models,providing reliable technical support for real-time system regulation,optimized decision-making,and subsequent development in smart water management initiatives.