Study Of Sparse Meteorological Data Processing Algorithm And Data Visualization Based On Participatory Sensing

Recently, traditional distributing platforms of meteorological information mainly include desktop clients based on C/S structure and publishing websites based on B/S structure, at home and abroad. And the main data source of these platforms is automatic meteorological stations. But influenced by many factors such as the number of automatic meteorological stations, space environement and communiation equality, there are sparse real-time meteorological data of automatic meteorological stations. If not handled in time, it will affect the continuity and real-time availability of the meteorological data set. According to the sparseness of traditional meteorological data, this thesis elaborates the design and implementation of a sparse meteorological data processing and visualization platform based on Participatory Sensing Meteorological Data. It makes use of Participatory Sensing technology and meteorological data of Anhui Meteorological Bureau. The main research contents and innovations are summarized as follows:(1) In view of the existing problems of the original RBF Neural Network, an RBF Neural Network with improved K-means is proposed. The improved RBF Neural Network is trained and learnt with meteorological data from automatic meteorological stations and mobile intelligent terminal. Simulation results show that the algorithm has high interpolation accuracy and strong reliability, and solves the sparse problem of meteorological data effectively.(2) This thesis introduces the application scenario of the platform. According to the actual requirements, the design goal is formulated. The general design framework,the function modules and the database design of this system are analyzed. Finally, the overall design of sparse meteorological data processing and visualization platform based on Participatory Sensing is given.(3) The meteorological data from automatic meteorological stations is analyzed.It introduces the process of collecting and uploading Participatory Sensing Meteorological Data. At last, the meteorological data visualization platform based on Spring MVC is developed. What is more, it provides a meteorological data upload and download module, a real-time meteorological data query module and a meteorological data statistics module.(4) According to the sparse meteorological data processing and visualization platform based on Participatory Sensing, the deployment of the server, the configuration of the platform environment and application testing are conducted.According to test analysis, the system is strong fault-tolerant, stable and scalable,achieves the expected research and design goal on the whole.