Case-based Reasoning For The Regional Metallogenic Prediction Method Study

Case-based reasoning is a kind of method which takes advantage of previous experience. Using analogy and association, it solves the current similar problems. And it is characterized by some advantages, such as low difficulty of implementation, high efficiency and accuracy, nonutility of deterministic regional model and easiness of settling uncertain problems, and so on.Mineral resource is a kind of non-renewable natural resource, with the features of hidden production, uncertainty of cognition and risk of reconnaissance. The fast development of modern industry results in the shortage of mineral resources. It is more and more difficult for searching minerals so that an intelligent method of seeking minerals has been developed to replace the direct geology reconnaissance. The intelligent method mainly predicts the location and production of minerals to reach the aim of finding minerals. In view of that background, the case-based reasoning was applied into the prediction of minerals in this study. The preliminary construction of the predicted method flow was made and the certified experiment was performed.Main work of the paper is as following:(1)Main methodology and techniques of the case-based reasoning including the eigen vector method of case-based reasoning and feature tree case-based reasoning method has been analyzed. Focusing on the key theories and technologies in the case representation, case retrieval, weight distribution, including feature tree method, analytic hierarchy process and fuzzy similarity methods et al, case-based reasoning method is represented.(2)The case-based reasoning method based on the eigenvector and the feature tree method are expanded. The case-based reasoning method fitting the demand of regional mineral resources prediction and experimental process is constructed.(3)Taking the regional mineral resources of east Kunlun and north marginal area in Qaidam Basin for example, basing on multi-source geological remote sensing data and the case-based reasoning method, experiment of the regional mineral resources prediction is completed. According to field data, the experiment is verified. Finally, use C++ programming language write a case based reasoning software system under the Microsoft Visual Studio 2005 platform.The experimental results indicate that the case based reasoning method which is used in regional mineral resources prediction forecasting is effective and feasible. Efficiency and accuracy of the regional mineral resources prediction is increased by the method. And the cost of finding mineral resources is reduced. And the method provided methodological supporting for the mineral exploration new again in our country.