| In recent years,with the vigorous and rapid development of China's society and economy,both the national development strategy perspective and the people's personal life have shown that the demand for mineral resources continues to increase.The overall level of mineral resources protection in China shows that supply is in short supply.How to deeply mine the inherent laws of data based on the existing geological information data,and discover the hidden information contained in it,has become an important research direction of current mineralization prediction and evaluation work.As a premise of the above research,collecting,collating,utilizing and sharing existing data has become a basic,long-term and important research task.How to use the information network platform technology to establish a data sharing service system is the basic condition for geologists and the public to further dig deep into the data law to find new breakthroughs.Based on the “National Mineral Resources Potential Evaluation” project,this paper takes the massive data of the formation of the national mineral resources potential evaluation process as the research basis,and through the combing and analysis of the results data model,proposes the B/S mode.The design plan of the national mineral resources potential evaluation results data management sharing system;and the system development based on the open source WebGIS platform,realizing the functions of rapid retrieval,scientific calculation,visualization,sharing and release of potential rating results data;On the basis of the management and sharing system,Hunan Xianghualing area was selected as the research area.With the rise of artificial intelligence in the era of big data,the deep learning method was innovatively introduced into the mineralization prediction work,and achieved certain results.The research content of this paper provides auxiliary tools for data integration of potential evaluation results,mineralization prospects,mineral resources survey and evaluation,etc.It also provides data support for further deepening the deep hidden information of potential evaluation results data,and proposes a basis for A new approach to predictive evaluation of mineral resources for deep learning.The main research contents and achievements of this paper include:(1)Potential evaluation result data has the characteristics of massive,multi-source and heterogeneous,which is a kind of big data.The traditional storage structure does not match the demand of big data.(2)On the basis of the metadata of the potential evaluation results,the metadata database of the result data is designed and constructed for unified management,so as to better store all kinds of data and provide support for the rapid browsing,query and retrieval of the subsequent data.(3)Design and develop a shared service publishing system based on the potential evaluation data of the open source platform B/S mode,which realizes common functions such as basemap switching,2/3D view switching,distance measurement,area measurement,and visualization of different professional data.(4)On the basis of the above,in accordance with the OGC general standard,combined with MapGIS IGServer,the WMS and other map services are released to realize the visual browsing and attribute query functions of various regions,themes,and minerals based on the three-dimensional virtual earth.(5)The Hunan Xianghualing area was selected as the research area through the system platform.The deep learning algorithm was used to predict the minerals in the area for the first time.Through the learning and training of the convolutional neural network model of the geology and geochemical data of the area,the verification results show that the model has certain Reliability,the automation and intelligence of the forecasting evaluation process are initially realized,and the stability and accuracy of the model still need to be further improved and optimized. |
PDF Full Text Request