In view of the difficulties in troubleshooting of high flexibility production line, a fault diagnosis method for power lithium battery module production line based on fault tree analysis and Bayesian network is proposed. Firstly, the mechanical structure and process flow of the production line have been analyzed. Combined with the collected fault cases, the fault tree of the automatic feeding station has been constructed to conduct qualitative analysis of the fault; Secondly, Netica software has been used to convert the fault tree model into Bayesian network model for quantitative analysis of faults; Finally, the actual fault cases of the enterprises have been substituted into the fault tree for verification and comparison. The verification results show that the accuracy of the fault diagnosis method combined with fault tree analysis and Bayesian network is 91.04%, which can realize the fault diagnosis of the whole production line, and the continuous iterative improvement of Bayesian network model in the diagnosis process can provide a basis for the improvement of subsequent production processes.