Regulation concept for web spreading equipment in web processing machines on the basis of neural networks

Web spreading systems are used in converting machines in order to guarantee a folding free web transport and a sufficient slit gap or slit separation. The adjustments, dependent on the respective application, are carried out manual before and during the operation. Wrong machine adjustments lead to productivity and quality losses. On this background a general regulation concept for web spreading systems, which is based on neural controllers is presented in this paper. The regulation concept is used for the regulation of the slit gap formation and applicated on the example of a dual spreader. The basic component of the controller is a controlling strategy which contains a mathematically describable, idealized operating range. All other, only qualitatively describable influence variables are declared as fuzzy data and included into the data base of the neural controller as training sets. First results show that neural controllers are suitable for the automatized operation of web spreading systems and can be drafted and simulated with passable operating expense