城市内涝模拟中机理驱动、数据驱动及混合模型的对比研究与进展综述OA

[Objective]In response to the need for urban waterlogging management under the background of accelerating urbanization and frequent extreme rainstorms,this paper aims to address the insufficient summary and comparison of current waterlogging simulation models(mechanism-driven,data-driven,and hybrid models),and provide support for refined waterlogging simulation and planning decision-making.[Methods]The research status of mechanism-driven models,data-driven models,and hybrid models is systematically reviewed.The technical characteristics,applicable scenarios,advantages,and limitations of various models are compared and analyzed,with a focus on analyzing the coupling paths and application cases of hybrid models.[Results]Mechanism-driven models can accurately depict physical processes but have limitations such as low computational efficiency and strong data dependence;data-driven models can achieve rapid prediction but face problems such as weak physical interpretability and limited generalization ability;hybrid models,by integrating the advantages of the two types of models,have shown outstanding performance in improving simulation accuracy and efficiency,becoming an important direction of technical integration.[Conclusion]This paper clarifies the applicable boundaries and development potential of different models,provides a theoretical basis and method reference for refined urban waterlogging simulation and intelligent decision-making,and highlights the practical value of technical integration in dealing with complex waterlogging scenarios.