水科学与水工程2008，Vol.1Issue(3)：73-82,10.DOI:10.3882/j.issn.1674-2370.2008.03.008
An information-based rough set approach to critical engineering factor identification
An information-based rough set approach to critical engineering factor identification
摘要
Abstract
In order to analyze the main critical engineering factors, an information-based rough set approach that considers conditional information entropy as a measurement of information has been developed. An algorithm for continuous attribute discretization based on conditional information entropy and an algorithm for rule extraction considering the supports of rules are proposed. The initial decision system is established by collecting enough monitoring data. Then, the continuous attributes are discretized, and the condition attributes are reduced. Finally, the rules that indicate the action law of the main factors are extracted and the results are explained. By applying this approach to a crack in an arch gravity dam, it can be concluded that the water level and the temperature are the main factors affecting the crack opening, and there is a negative correlation between the crack opening and the temperature. This conclusion corresponds with the observation that cracks in most concrete dams are influenced mainly by water level and temperature, and the influence of temperature is more evident.关键词
rough set/information entropy/critical engineering factor, crack openingKey words
rough set/information entropy/critical engineering factor, crack opening引用本文复制引用
Jing Ji,Zheng Dongjian..An information-based rough set approach to critical engineering factor identification [J].水科学与水工程,2008,1(3):73-82,10.基金项目
This work was supported by the National Natural Science Foundation of China (Grant No.50539010 and 50579010) and the Ertan Hydropower Development Company Joint Research Project (Grant No.50539030-1-3). （Grant No.50539010 and 50579010）
