Model based self-optimization of the weaving process

Abstract Warp tension is a critical variable of the weaving process. If the warp tension is too high or too low the weaving process will be interrupted. In order to find suitable setting for the weaving machine, the experience of the operator or data base systems are used. Within this paper an automatic setup routine following model based self-optimization strategies is proposed. Within the routine, data for a regression model are collected by the weaving machine. For given quality criteria the weaving machine is able to calculate an optimal setting point. Validation of the routine shows that the chosen regression model is suitable; stress on the warp yarns is reduced. In addition, a statistical validation proves the usability of the regression models.