Model Predictive Control of the Weft Insertion in Air-jet Weaving

Abstract

Abstract During the weft insertion in air-jet weaving, a yarn brake plays an important role in reducing the tension when braking the weft thread. An inappropriate braking process can cause backward movements of weft thread, fabric defects and also larger material consumption. For these purposes, this paper proposes a model predictive control system. Regarding external forces from nozzles and weft internal dynamics, a weft insertion process model is developed. Particularly, as the essential accelerating force, the aerodynamic force is dependent on the dynamics of airflow in time and space domain. The model is then validated in order to determine the model parameters. Considering state observation and computational efforts, a control-oriented model is derived based on the analysis of system dynamics. A model predictive controller takes advantage of this reduced model to predict the system behavior in the future. Furthermore, the constraints on the velocity of weft tip are considered explicitly in the optimization problem such that backward movements of the weft thread are avoided. Finally, the simulated position and velocity behaviors of weft thread are presented. The controlled weft is able to arrive at the desired position without constraints on velocity violated.