Application of adaptive neuro-fuzzy system for tape rectification

The scheme of adaptive control for tape rectification on a robotic assembly line is introduced. The control algorithm of the system is split into control stage and recognition stage. In control stage, the discrepancy between desired tape edges distance and actual tape edges distance is obtained by vision processing system. A forward multilayer fuzzy neural network performs the map from the discrepancy to command signal, which is then used to actuate rectifying mechanism, according to fuzzy rule inference. In recognition stage, the fuzzy model of controlled object is formed and the parameters of the fuzzy neural network are updated. The effectiveness of the adaptive system is demonstrated by experimental results.