| Tartalmi leírás: |
In recent years, rapid urbanization has significantly impacted water quality in urban areas. Accurate prediction models for pollutant dispersion play a pivotal role in enhancing our ability to respond effectively to pollution incidents and promote sustainable urban development. However, the computational cost brought by traditional simulation models is relatively high and time-consuming. To address this problem, surrogate models are employed to enable efficient and rapid evaluations. This paper proposes a simulation-based prediction method for river water quality. First, a one-dimensional hydrodynamic water quality model is established by HEC-RAS. The riverbed data is derived through cubic spline difference, facilitating the simulation of the water pollutant transport process. Secondly, the hydraulic and water quality simulation results are obtained from the model constructed above. Subsequently, the simulation results are used as input to a deep neural network for further training and learning processes. By evaluating the applicability of deep learning models in predicting water pollution, this method can provide a valuable reference for urban water governance.
|
