The rivet assembly is an essential work for the aircraft manufacturing. Its quality is highly related with the flight safety. This paper presents an intelligent quality inspection system which is constructed by a fuzzy mechanism and a neural network (NN) for the riveting work. The fuzzy mechanism plays the role of signal processor which is used to enhance the significant features of the impact signals so that the neural network could make the immediate diagnosis of the riveting quality based on the processed signals. The intelligent inspection system is expected to give the assistance for the verification of rivet quality. It can not only monitor the real time riveting process, but also greatly decrease the inspection inaccuracy caused by the visual examination. The experimental data provided by Chinese Air Force Institute of Technology were simulated and studied. In order to demonstrate the superiority of the intelligent inspection system developed, the pure neural network model with no fuzzy mechanism was also performed as the comparison.

Full text: IJEIR_782_Final.pdf