A neural network-based assembly algorithm for chamferless parts mating

This paper presents a neural network-based assembly algorithm for chamferless parts mating. The difficulties in assembly strategies result from the errors of assembly system as well as the complexity of the assembly process. To solve this problem, we propose an active-passive assembly algorithm by integrating a passive vibration method and an active force feedback signal, This method generates a vibration pattern by signs of moments, which is used to compensate the misalignment between mating parts. In this method, the relationship between the signs of moments and output patterns of the vibration is trained by neural network. The performance of the presented method is evaluated through a series of experiments. The experimental results show that the method can be applied to chamferless parts mating