| Digital Park is part of the mandatory fields related to Digital Earth and is also a transitional phase for the development of Smart Park.The application of Digital Earth involves many fields of geophysical prospecting,including a series of scientific issues such as geophysical data processing,geophysical data mining learning.In recent years,the research interests of a smart park have increasingly aroused in the world,especially on the critical technologies.Such as the construction of the smart park platform,big data storage,and analysis,application of intelligent algorithms,the network architecture of related algorithms,combined with the GIS applications,etc.Digital Park covers digital mines.With the continuous deepening of prospecting,the number of discoveries and exploitable mines in the shallow areas has been reduced.Research on mineral exploration and data storage computing platforms involved in deeper pyrogeophysical prospecting belongs to the research area of digital parks The development of mineral resources prediction and evaluation is intimately linked with the mining exploration process.Initial mineral prediction was mainly based on geological conditions.Since the 1970 s,with the rise of scientific prospecting and the extensive and mature application of prospecting technologies such as aviation,remote sensing(RS),geophysics and geochemistry,mineral resources prediction It has entered the new phase.Finite element,discrete element,boundary element and finite difference model are the research tools of modern digital mining area.The finite element model is used to study the physical exploration,and the data collection Analysis and processing by the digital campus platform is a big part of data processing.Data-Oriented Architecture(DOA)based on relevant cloud technology,reflecting the idea of “data-oriented and data as the core factor.” This dissertation applies DOA theory to construct the platform of the smart park,collect data from Digital Park,to solving some scientific problems about the development of the digital park to smart park,it also innovatively puts forwards the architecture based on DOA theory,and build a platform model for high-performance computing and mass data storage through independent development and open source software code development.Smart Park takes G/S(General-Brower/Distribute Spatial Information Server)model to manage and show the views via the development of Qgis,PostGIS,and OpenLayers,etc.For the question about the insensitive performance of machine learning on location,in the ML(Machine Learning,ML)method of big data and various structural/nonstructural data,aiming at the problem that the original DeepLearning.SVM,RF and the like are insensitive to the geographical location,the concept of geographic location parameter of GIS is introduced,this dissertation creatively puts forth the location-based machine learning(LBML)algorithm and proves its validity through the experiments on data correlation extraction and location-based data learning,etc.For the combination of cluster analysis?classification and prediction algorithm,this dissertation also puts forth the multilayer machine learning network(MLMLN)model and proves its superior performance on self-organizing a multi-layer network of machine learning algorithm,The model constructs a self-organizing multi-layer machine learning network in the process of solving the complicated and complicated intelligent requirements.The model forms a network and automates the processing of data,and optimizes the route selection from data input to data output The holistic platform is utilized to digital park and the related fields of using Big data for the location-oriented geophysical prospecting,Dataoriented smart park platform in all kinds of heterogeneous data in digital parks and geophysical exploration in digital mining area of various types of experimental data collection and classification experiments,the results show that the platform has a good effect... |
PDF Full Text Request