Deep deterministic policy gradient based state feedback control for warp tension in fishing net weaving machine

Cankun Ming,Yujing Zhang,Zhijun Sun,Yize Sun
DOI: https://doi.org/10.1080/00405000.2024.2361173
IF: 1.7698
2024-06-27
Journal of the Textile Institute
Abstract:The warp tension of the fishing net weaving machine significantly influences its weaving efficiency and mesh quality. Traditional fishing net weaving machines control the warp tension using the self-gravity of the knot-tightening arm, which is imprecise and cannot suppress external disturbances. In this study, we proposed a knot-tightening mechanism with servo motors and a state feedback controller to regulate warp tension by adjusting the torques of the servo motors. Additionally, we incorporate a deep deterministic policy gradient algorithm to dynamically adjust the controller parameters. Simulations and experimentations suggest that the deep deterministic policy gradient based state feedback controller outperforms both the conventional PID controller and the state feedback controller with fixed parameters in terms of both control precision and disturbance resistance capability.
materials science, textiles
What problem does this paper attempt to address?