VIBRATORY ASSEMBLY OF PRISMATIC PARTS USING NEURAL-NETWORK-BASED POSITIONING ERROR ESTIMATION

Despite its known effectiveness, a typical vibratory assembly method t ends to generate adverse impact forces between mating parts commensura te with the relatively large vibratory motion required for reliably co mpensating positioning errors of arbitrary magnitude. To this end, thi s paper presents a neural network-based vibratory assembly method with its emphasis on reducing the mating forces for chamferless prismatic parts. In this method, the interactive force is effectively suppressed by reducing the amplitude of vibratory motion, while the greater part of the relative positioning error is estimated and compensated by a n eural network. The estimation performance of the neural network and th e overall performance of the assembly method are evaluated experimenta lly. Experimental results show that the assembly is efficiently accomp lished with small reaction forces, and the possible insertion error ra nge is also expanded.