The key to realize intelligent manufacturing cell is the timely response of CNC machine fault,accurate judgment of fault causes and rapid identification of solutions.Based on the historical operation and maintenance data of CNC machine tools,association rules mining FP-Growth algorithm was adopted,and association rules mining of fault modes and causes of CNC machine tools was realized.The fault characteristics of some historical operation and maintenance data were analyzed.A fault diagnosis model based on association rules was constructed.By mining the historical fault modes and corresponding fault causes,the specific fault diagnosis association rules were obtained.Finally,the feasibility and rationality of the FP-Growth algorithm for fault feature analysis of CNC machine tools were verified by analyzing the evaluation indexes such as support degree and confidence degree of association rules,and deducing the magnitude of association probability.