Performance of dissimilar materials riveted joints based on fuzzy RBF neural networks
                                                                               CHEN Fang, LI Changsheng
                                (School of Mechanical Engineering, Henan Institute of Technology, Xinxiang 453002, China)

Abstract: Aimed at the problems of low reliability and poor connectivity for dissimilar materials riveted joints, the effects of the process parameters such as the rivet size and punch stroke on the mechanical properties of riveted joints were investigated by the method of fuzzy RBF neural network. The fuzzy RBF neural network model was introduced to establish the mapping relationships between structure parameters of the rivet size, punch stroke and strength parameters such as shearing force and peeling stress, and to predict the mechanical properties of riveted joints. The results indicate that compared with common BP neural network, the relative prediction error of shear strength and peel strength are greatly reduced by the fuzzy RBF neural network model. The fuzzy RBF neural network is introduced into the model that can accurately analyze the relationships between process parameters and mechanical properties of the riveted joints.

Key words: dissimilar materials; riveted joint; mechanical properties; fuzzy RBF neural networks