Backstepping control with radial basis function neural network for web transport systems

Abstract

Web transport system (WTS) is commonly used in the production and handling of web materials such as paper, fabric, corrugated iron, steel, and printing operations. These materials are easily damaged if the process performance is poor. Therefore, high technology in mechanics and precise control techniques are required in this system. In addition, because of parameter variation, strong nonlinearity, and many external noises, there are many challenges in controlling this process. This paper proposes a backstepping technique-based algorithm to control the web’s tension and velocity. To solve the parameter variation, a radial basis function (RBF) neural network-based adaptation algorithm is developed to approximate the varied components in the control algorithm. The system stability is guaranteed using the Lyapunov stability theorem. Simulations in MATLAB/Simulink have been done and the effectiveness of the proposed control algorithm is verified. Tension and velocity tracking can be obtained with parameter variations.