| In recent years,the frequent occurrence of extreme weather makes the economic losses caused by mountain torrents increasing day by day,and even causes a large number of casualties.For this reason,the state has invested a lot of manpower and material resources to carry out the prevention of mountain torrents,and has achieved fruitful research results,but mountain torrents still occur from time to time.Compared with the flood control and disaster mitigation decision-making of large rivers,mountain flood disaster prevention needs fast and accurate prediction and early warning information in order to avoid danger and transfer in advance and achieve the purpose of disaster mitigation.At present,mountain torrent forecasting and early warning is mainly based on the critical rainfall released by the monitoring and early warning platform at County level.Due to the influence of underlying surface,topography,rainfall intensity and spatial distribution,strong rainfall does not necessarily cause disasters,and often there are omissions and false alarms,which bring difficulties to the prevention of mountain torrent disasters.Therefore,in order to improve the level of mountain flood forecasting and early warning,it is necessary to establish a multi-factor and multi-mode early warning integrated platform.In view of the above problems,this paper takes the project of monitoring and early warning of mountain torrent disasters in Liaoning Province as the background,and carries out the key technology research on the platform of mountain torrent disaster prediction and early warning.Firstly,the structure characteristics of hydrological model are analyzed,and the integrated method of hydrological model is studied by computer technology;secondly,the applicability of hydrological model is analyzed,the index system of hydrological model identification is established,and the application of hydrological model identification is studied;secondly,the multi-dimensional forecasting and early warning model is developed according to the confluence characteristics of different basins and the practice of mountain flood disaster prevention and early warning;finally,the research will be completed.The results are applied to Liaoning Province's mountain torrent disaster forecasting and early warning platform system to realize multi-factor,multi-scale,multi-mode and multi-level forecasting and early warning platform,and to improve the level of mountain torrent forecasting and early warning.The main research contents and achievements are as follows:(1)The research status of mountain torrent disaster prevention and control at home and abroad is analyzed,the existing problems of mountain torrent early warning in China are pointed out,and the research content of this paper is clarified.(2)According to the actual needs of mountain torrent forecasting and early warning,the research of multi-model hydrological model integration method is carried out.Firstly,the mechanism and structure characteristics of runoff generation and confluence of main watershed hydrological models are analyzed,the calculation process of watershed hydrological models is clarified,and watershed hydrological models are divided into independent calculation modules;secondly,business calculation components of interface standardization are developed by using object-oriented technology,database technology and metadata technology;secondly,appropriate components are selected and integrated according to actual needs.Establish different flood calculation models.The application results show that the model integration method improves the system development efficiency and code reuse rate,and facilitates the integration of hydrological models under different application conditions.(3)In order to adapt to the identification of hydrological models under different application conditions,the mechanism of runoff yield and confluence and the applicable conditions of the main basin hydrological models are analyzed.Based on this,the main factors affecting the selection of hydrological models are screened by using principal component analysis,considering hydrological zoning,basin characteristics,temporal and spatial distribution,underlying surface conditions and so on,and basin water is established.The intelligent identification index system of hydrological model in this paper can be used to identify and select hydrological model parameters calibration or flood forecasting operation suitable for current conditions according to the intelligent identification index system of hydrological model in river basin established.(4)In order to solve the problems of false alarm and missed alarm of single critical rainfall index in mountain torrent early warning,this paper carries out rainfall-based early warning model,pseudo-forecast and disaster-causing flow-based early warning model,real-time forecast and disaster-generating flow-based early warning model from the underlying surface conditions,spatial and temporal distribution of rainfall,real-time rainfall,pseudo-rainfall and other factors and different time scales of 1 hour,2 hours and 3 hours.Quantitative early warning model and other early warning models are studied and verified by examples.The results show that the proposed multi-dimensional forecasting and early warning model improves the effectiveness and reliability of early warning.(5)Based on Multi-model hydrological model integration and intelligent identification method of model,multi-dimensional forecasting and early warning mode,using J2 EE technology,MVC design mode and Spring system framework,the mountain flood disaster forecasting and early warning platform in Liaoning Province is designed and developed.This early warning platform can realize the functions of intelligent flood forecasting,multi-dimensional flood forecasting and information query,which greatly improves the validity and reliability of mountain flood disaster early warning. |
PDF Full Text Request