Fault diagnosis of water chiller based on improved Apriori association rule mining

Liu Qinggui and Ding Jinliang

2018.04.11

For the unpredictable problems of water chiller faults in normal operation, proposes a fault diagnosis method with association rule mining. Based on the history database of the water chillers, figures out the relationship between the variables and the faults. Proposes an improved Apriori algorithm, whose support can make the association rule mining results meet the requirements. The results show that the updated algorithm significantly improves the accuracy, the number of scanned database and the running time.