Aiming at the problems of complex failure of the artillery filling system and the lack of diagnostic methods, a method of gun loading monitoring and fault diagnosis based on multi-source information fusion is proposed. The knowledge decision attribute is used to classify the attributes, and the neural network training model is constructed. The faults of the automatic loading system are qualitatively analyzed and the fault diagnosis identification framework is established. The fault judgment is obtained according to the fault decision criteria, and the uncertainty of the fault features and the varied fault modes are solved. The example analysis shows that the method achieves the goal of effectively improving the diagnosis rate of fault diagnosis.