In view of existing problems at current aquaculture networking,such as nonlinear characteristic of fault and faults are easily affected by many factors,a fault diagnosis model based on neural network was proposed. In the building process of the model,the common fault types in the field of aquaculture networking were first analyzed and the types of fault mode were summarized. Afterwards,the evaluation indices of fault diagnosis were made,and eventually the fault diagnosis system of aquaculture networking was constructed using neural network principle. The fault diagnosis system could not only reduce the communication burden,but also have high diagnostic rate. Thus,it could be well applied in the fault diagnosis system for aquaculture networking. In view of existing problems at current aquaculture networking, such as nonlinear characteristic of fault and faults are easily affected by many factors, a fault diagnosis model based on neural network was proposed. In the building process of the model, the common fault types in the After of the evaluation indices of fault diagnosis were made, and eventually the fault diagnosis system of aquaculture networking was constructed using neural network principle. The fault diagnosis system could not only reduce the communication burden, but also have high diagnostic rate. Thus, it could be well applied in the fault diagnosis system for aquaculture networking.