Application of Data Mining in Coal Mine Safety Decision System Based on Rough Set

Abstract

The coal mine safety decision systems, such as ventilation safety monitoring system, underground water inrush monitoring system, underground coal and gas emission monitoring system, have been established in many large and medium-sized coal mines. A large amount of original data had accumulated in these systems. How to transform data into information for scientific decision was a problem worth to consider for coal mine safety production. The rough set theory, quantitative analysis of incomplete, imprecise and uncertainty knowledge, provided a new method and tool for data mining. A kind of heuristic genetic algorithm for continuous attributes discretization was put forward to solve the problem of continuous attribute discretization of decision table; a kind of heuristic immune algorithm for attribute reduction was presented to conquer the shortage of existing attribute reduction algorithm; in order to solve the problem of reasoning and decision in incomplete and imprecise information, a kind of default rule mining model based on reduction lattice was proposed. Finally, data mining system based on rough set was designed, which was applied to data mining analysis of underground gas emission, good results were achieved.