The Model Construction And Application Effect Of Anomaly Threshold Extraction Based On Genetic Algorithm And Spatial Interpolation Algorithm

The research area of this thesis is Xishanwanyangchang silver deposit in Inner Mongolia,which is located in the eastern section of the Langshan-Zha'ertaishan Mesoproterozoic to Neoproterozoic age rift zone in the northern margin of the North China platform.The structure in the study area is relatively developed and geological conditions are beneficial to mineralization.Anomaly threshold is a research focus in exploration geochemistry.The related research results are numerous.On the basis of reading a large number of literatures,documents and papers,this thesis collected some technical methods related to anomaly threshold in recent thirty years.And its operation flow and characteristics are briefly introduced.In this thesis,the research ideas of anomaly threshold are summarized as two aspects.One focuses on data de-noising,and the other focuses on distinguishing the background information and abnormal information from data.Based on a series of induction,a new model is proposed to solve the problem about anomaly threshold.The model consists of two sub-models,namely the practical algorithm and the comprehensive model.The practical algorithm is stripped from the comprehensive model.The practical algorithm integrates the knowledge of spatial interpolation,numerical simulation and probability theory,therefore,this method is named as spatial probabilistic simulation(SPS).The comprehensive model integrates the knowledge of genetic algorithm,spatial interpolation principle and the central limit theorem,which is called the GA-SI-CLT model.This thesis uses Matlab to realize above models.Through the application of the stepwise elimination method,the C-A fractal method,the practical algorithm and the comprehensive model,the value of anomaly threshold can be obtained.On the basis of comparison and analysis,it is concluded that the new model proposed in this thesis has certain feasibility and application potential.In addition,the comprehensive model can identify weak anomaly sensitively and the practical algorithm has the characteristics of high efficiency and can outstanding anomaly.Therefore,these models can take into account the efficiency and effectiveness of abnormal delineation.