| In the 21st century, social and economic development is inseparable from resources, and the demand for mineral resources is increasing among various kinds of resources. It has become an important issue of fast and accurate prospecting. Although the ore deposit location prediction develops continuously, there are a lot of new challenges. With the standardization, institutionalization of mineral prediction in China and the variety of geological information data increasing, the GIS which represents spatial information science has been widely applied in mineral prediction and got great development. This paper analyzes the domestic and foreign mineral prediction theory and methods, elaborates modern spatial information science applications in mineral forecasting, and combines both to process and predict mineral information with cloud model algorithm.People get a great deal of earth physics, chemistry and other related information from earth exploration technology permanent development, but software research and data analysis techniques in spatial information science such as remote sensing and GIS move slowly, leading to a significant portion of the data is not made full use of, and unable to dig up valuable and meaningful knowledge. At present, the spatial data mining technology is developing rapidly, generates a lot of new technologies and data mining algorithms model. It's easy to use spatial information technology like GIS software to collect, store, edit, manage and analyze data under the combination of spatial data mining research, spatial information technology and regional mineral resource prediction.Cloud model is merging both the probability theory and fuzzy mathematics and it's developing. It is also an uncertainty transformation model between qualitative concept and quantitative data. There are many qualitative concepts in human social life, and it is not directly expressed with data, whereas quantitative data can't be converted into good qualitative concept. How to better cany out conversion of both is an important issue. However, the cloud model is a good solution to this problem with the fuzziness and randomness. Mineral prediction also has a lot of uncertain information, and it's accompanied by randomness and fuzziness. In the process of uncertainty research, we can convert the complicated social language qualitative concept to simple data. This paper describes the one-dimensional and two-dimensional cloud model concepts, algorithms, and draws the image by Matlab. It predicts metal mineral reserves in Tongling mining area. Then the paper researches uncertain information fusion in mineral prediction of, and predict unknown mineral. The main work is as follows:(1) Since geological data sources is more diverse and has errors and deficiencies during Tongling data preprocessing, so it corrects and removed the wrong place by the pretreatment. Ensure data consistency, accuracy and completeness. This study is based on Anhui gravitational fields, magnetic, geochemical, geological, mineral reserves data and fault information mainly through ArcGIS, MapGIS software editing.(2) Introduced the concept of cloud model, algorithm and uncertainty reasoning. At present, there are one and two dimensional cloud model, and describe the the two cloud model numeral characteristics, cloud model morphological characteristics, cloud forward generator and cloud reverse generator algorithm, and the cloud model uncertainty reasoning model. Use Matlab language programming to draw one and two dimensional normal cloud model.(3) After the data of Tongling city acquisition and pre-processing, the metal mineral reserves reasoning model is established and convert complex, chaotic geological survey data into a cloud model, and build multi-rule reasoning mode on the base of cloud model. Then predict metal mineral reserves in Tongling City and surrounding area. Through the comparing between predicting and actual data it indicate the prediction is good. |
PDF Full Text Request