Weighted adaptive generalized predictive control for carding autoleveler

In textile industry, carding process has decisive influence on produced yarn quality. From the system theoretic point of view, it is marked by stochastic disturbance, long delays, and parameter variation. So, a carding process is difficult to control with traditional control algorithms (such as PID). In this paper, a weighted adaptive generalized predictive control (GPC) law was developed to control such a process. The experimental results show that GPC autoleveller controller could greatly reduce the sliver's standard deviation and reject disturbance.