基于粗糙决策模型和蚁群算法的故障诊断

Fault Diagnosis based on Rough Sets Decision Tree Model and Ant Colony Algorithm

粗糙集理论是一种新型的处理模糊和不确定知识的数学工具,它能在保持信息系统分类能力不变的前提下,有效地进行知识约简;决策树算法可对约简后的决策表进行规则提取,有着直观、能定量分析的优点.提出一种变化趋势关联度来评价条件属性的重要性,建立粗糙决策树故障诊断模型,并在诊断推理中引入一种新的人工智能方法——蚁群算法,来确定决策树的最优检测次序.将其应用于工业精对苯二甲酸(purified terephthalic acid,PTA)生产过程中的对二甲苯氧化反应温度的诊断结果表明了提出方法的有效性.

Rough sots theory is a new mathematical tool to deal with vagueness and uncertain, which can remove redundant information and seek for reduced decision tables effectively. Decision tree can extract diagnosis knowledge from reduced decision tables in the form of symbolic trees and easily understood inference mechanisms. Trend correlation degree, as the heuristic knowledge, is proposod to evaluate the significance of condition attributes for the construction of decision tree model. Also a new intelligent algorithm called ant colony algorithm is introduced. With its outstanding characters, the optimal test sequence can be achieved. The proposed method is applied to the fault diagnosis of reaction temperature in industrial purified terephthalic acid （PTA） oxidation process. The effectiveness of the method is therefore proved through the exemplification.