Development of hysteresis friction model for a precise tracking control system

Hysteresis friction model for a tracking control system of the mechanical system was proposed in this article. Hysteresis phenomena within a microscopic displacement range were verified through experiments, and Preisach model was applied to the proposed friction model. To implement the Preisach model, the neural network algorithm was used to replace the distribute function of Preisach model. A weighing vector of the neural network was updated through the back-propagation learning method. The performance of the proposed hysteresis friction model was evaluated through the tracking control of a ball-screw system. In order to design the tracking control system, sliding mode controller was designed and the proposed friction model was applied as a feed-forward term of controller.