Study On Rough Set Theory In The Prediction Of Coal Reservoir Permeability

Coal Reservoir permeability is an important parameter of estimating favorable exploration target area. However, there are only common methods and little data in the coal exploration area. Rough Set theory works as an effective way of dealing with uncertainty and imprecise data. It works well during the process of Data Mining. it is used sucessfuly in all fields. In this paper the rough set theory is used in the predition of Coal Reservoirs Permeability and provides a new methed for it.Firstly, this paper summarizes the rough set theory and expounds some important concepts, such as knowledge, indiscernibility relation, attribute importance and so on. it outlines the regional geology. This paper centres on the main effect factors including effective stress, cleat, buried depth and so on. After that it chooses suitable effect factors as impact factors. The process of establishing the model is based on the rough set theory consists of assumptions, the choice of indicators data, preprocessing, attribute reduction, rule creation and rule testing. During the study, the program realized the main algorithms of the Rough Set theory utilizing Visual C++6.0.This thesis use modles in southern Qinshui basin to do empirical study. And Select discernibility matrix attribute reduction algorithm and heuristic value reduction algorithm for processing the decision table. The results show that the modle of the predition of Coal Reservoir permeability which is based on the rough set theory has great predictive ability. The accuracy rate is 90% which shows that the model has good predictive ability. We use the rules to predict permeability distribution in the study area. Permeability distribution permeability distribution of the study area shows that the south area is higher than the north area. The permeability of Jincheng is generally greater than 1×10^-3μm², it is concluded that Jincheng area is favorable zones for exploration.knowledge reduction is the core of Rough set theory, through the reduction of knowledge we can found the rules, in the establishment of forecasting model, we can simplify sample of rough set and extract the core attributes.