Aimed at the problems that structure and function of aero-engine are complex, construction of Bayesian network is difficult, and it is difficult to obtain the exact value of node conditional probability, in this paper, a knowledge graph with fuzzy Bayesian network (KG-FBN) inference fault diagnosis method is proposed. Firstly, on the basis of large-scale historical fault data, an aero-engine fault knowledge graph is constructed by using the knowledge graph technology. Secondly, a mapping method of “knowledge graph-Bayesian network” is proposed to rapidly construct Bayesian network, and introduce fuzzy set theory to solve the uncertainty problem of probability parameters in engineering practice. Finally, an example is given to verify the feasibility of the proposed method. The results show that the proposed method can improve the efficiency of Bayesian network construction and achieve uncertain inference in fault diagnosis, can be also used for optimizing diagnostic strategies, and can improve equipment reliability, and is strong in engineering application value.