Fault diagnosis of industrial boiler based on competitive agglomeration and fuzzy association rules
Hui Juan Zhao, Jiang Bi-bo, Zhuoqun Zhao · 2010
Applying datamining algorithm to the association analyzes between the measurable parameters and faults in the industrial boiler control system. The works we have done generally as follows. According to the distribution of the parameters that can be measured, using competitive agglomeration clustering algorithm to partition the fuzzy interval of each attribute; based on the principle of association rules, an algorithm has been proposed to find the association between the parameters that can be measured and the fault; The proposed algorithm has been realized and validated. As the results proves that fault diagnosis of industrial boiler based on competitive agglomeration and fuzzy association rules can mining knowledge effectively, and the knowledge has higher correct rate than normal methods. The project was supported by National High-tech R&D Program (863 Program) (2007AA041401), Tianjin Natural Science Foundation (08JCZDJC18600, 09JCZDJC23900), and University Science and Technology Development Foundation of Tianjin (2006ZD32).