Fault diagnosis based on ant colony algorithms

摘要

For concluding the difficulty of classing fault sign of equipment automatically in fault diagnosis, this paper presents a new excellent clustering algorithm based on Ant Colony Algorithms (ACA). It is discovered the diagnosis earlier, it is classified fault sign of equipment automatically, and obtain diagnosis knowledge, conclude diagnosis rule, find the reason of fault. All these are in favor of fast, automatic and exact decision-making and dealing with the fault. ACA is applied in the fault diagnosis and recognition, and does pattern recognition for a chemical reactor. Result and the actual operation state are consistent. That can reflect the algorithm accuracy.