Intelligent automatic alignment of fibre optic components

This paper presents an 'intelligent' automatic alignment system for use in the assembly of optical fibre components. The system optimises the light output from a fibre by applying a gradient-based algorithm enhanced with momentum information. The algorithm parameters are adjusted on-line by a fuzzy controller according to the progress of the alignment process. Using past experience, a neuron-like element directs the starting point of the search for the optimum alignment towards the most favourable position. The automatic alignment system yielded 100% success in a test involving 100 different samples of components. The system achieved significant reductions in alignment times compared with manual or conventional automatic methods. The system was also almost twice faster than an alternative based on Simulated Annealing. Moreover, the use of fuzzy logic facilitated the design of the controller and provided noise-rejection capabilities.