| The geological environment of Qin-Ba Mountain Area is extremely complex,and landslide disasters are particularly frequent under the influence of extreme climate and human engineering activities.With the increase of investment in landslide disaster monitoring and early warning by relevant departments,the monitoring data are also increasing exponentially.Driven by monitoring technology,the monitoring data are more complex and diversified,and the timeliness is increasing day by day.The management of massive monitoring data,the integration of multi-source heterogeneous data and the automatic extraction of dynamic data information are becoming increasingly prominent.Therefore,it is very important to construct a monitoring and early warning platform for efficient collection,processing,storage and integration of various monitoring data,and provide technical support and solutions for disaster prevention and mitigation in Qin-Ba Mountain Area.In this paper,Qin-Ba Mountain Area is taken as the research area,and landslide monitoring and early warning is taken as the research object.Based on big data and information technology,combined with landslide early warning model,a landslide disaster monitoring and early warning platform in Qin-Ba Mountain Area is constructed.The main research results of this paper are as follows:(1)According to the characteristics of landslide monitoring data in Qin-Ba Mountain Area,a complete data acquisition,processing,storage and integration system is designed.The platform uses the enterprise-level geographic database Postgre SQL and the distributed file system HDFS to complete the storage of structured and unstructured data.Finally,the platform integrates the data based on the middleware model based on Arc Py and Spark environment,forming an early warning platform data center with enterprise-level geographic database and distributed file system as the core.(2)Based on the disaster-causing factors of landslide disasters in the Qin-Ba Mountain Area,the artificial neural network model and the logistic regression model were used to evaluate the landslide susceptibility in the Qin-Ba Mountain Area.The susceptibility distribution map generated by the model with higher accuracy in the two results was selected as the spatial probability by means of quantitative methods.Coupling it with the probabilistic rainfall threshold model,a regional landslide warning model in the Qin-Ba Mountain Area was obtained.The infinite slope model and unsaturated soil stability analysis model are selected to be applied to the monitoring and early warning of watershed and landslide point respectively.According to the relevant research,the multi-scale landslide warnings such as regional,watershed and single slope are graded,the measures to be taken for each warning level are clarified,and the monitoring and early warning process of the early warning platform is designed.(3)Based on the functional and non-functional requirements of the platform,the overall design of the early warning platform is carried out from the aspects of overall architecture,technical selection,technical architecture and deployment design.On this basis,according to the needs of the platform and the data characteristics of the study area,the design of the early warning platform database,API and system security is completed.According to the module function of the platform,it is divided into five categories,and the realization of each functional module is completed.(4)Based on the design and implementation of the early warning platform,the early warning platform is tested.The test results show that the early warning platform basically meets the requirements in terms of function and performance.Finally,according to the example of landslide event,the landslide early warning model is used to calculate the early warning.The results show that the accuracy of regional early warning is high,and the basin early warning can further reduce the scope of key early warning areas on the basis of regional early warning. |
PDF Full Text Request