Improvement of Paper Quality Management System by Web Width on the Basis of Intellectual Technologies

Improving the quality of paper products on wide paper machines involves the implementation of high-precision control of the main indicators of paper quality, such as basis weight, moisture content, and thickness. The control of each of them consists of dozens of actuators with overlapping areas of operation, and thus poses the challenge of simultaneously coordinated control of the entire line of actuators. In addition, the basic weight, moisture content, and thickness of the paper web are interconnected through these lines of actuators, the control of which affects several quality indicators. This results in the problem of controlling a complex object with a large number of interrelated parameters. To solve this problem, this paper proposes to use a control system of several neural controllers united by a common criterion of reducing the total control error.