In order to solve the uncertainty and complexity of flight deck wheel fault diagnosis, a fault diagnosis method based on Bayesian Network was proposed. Firstly, the K-S test method was used to verify that the fault data obey the exponential distribution, and then the maximum likelihood estimation algorithm was used to optimize the parameters to obtain the optimal failure rate of basic events. Secondly, the Bayesian Network was constructed according to the failure tree of flight deck wheel. Finally, the failure rate was put into Bayesian Network model for analysis. The results show that the BN network model has high reliability and accuracy, which can provide strong support for the maintenance personnel to make quick judgment and efficient maintenance.