Modeling And Compensation For The Friction In Mechanical Motion System

Friction exists widely in electromechanical servo system. Friction, as a non-smooth andnon-linear phenomenon, is one of the main factors that lead to tracking errors, limit cycle, andstick-slip movement, etc. Therefore, we must consider the influences of friction at the design ofcontroller. Friction has been studied for lots of year, however, due to the complexity of frictionthere are urgent needs to address the problem in the theory and practice.Along with the development of friction models, the methods of friction compensation arepromoted forward and model-based friction compensation strategy has become mainstreams.Many friction model parameters have the characteristics of the slowly-varying uncertainty, inorder to deal with these problems, adaptive controller combined with the advantages of othercontrol strategies, which generate a class of combination controller, such as the robustadaptive control, adaptive sliding mode control, adaptive neural network control, fuzzyadaptive, etc.To achieve a good compensation of friction, friction modeling, identification andcompensation become a hotspot. This paper focuses on the friction impact to the control systemand how to reduce or even eliminate the negative effects of friction. The main content of thepaper is as follows:First, we summarize current researches and applications of the friction model by readingthe literature. Understanding of the friction model has gone through the stage of development ofthe static and dynamic models. In many engineering applications, the classic static model is stillwidely used due to the simple and practical, however, high-precision movement control areas,such as high-precision CNC machine tools, robot joint control and micro-electromechanicalequipment unit, the dynamic model has been developed greatly and used widely.Second, introduce the components of mechanical movement system (SRV-02) and create anonlinear model. By analysis the characteristics of the model, which is divided into the linearpart and nonlinear part, we identifies the parameters of Stribeck friction model and takePID-Feedforward strategy to validate the identification results.Finally, we design nonlinear PI adaptive controller by combine the advantages of thenonlinear PI controller and adaptive controller. Nonlinear PI controller has strong robustness, the adaptive controller can deal with parameter uncertainty, using the method combining twocontrol strategies improve the control effect significantly, system robustness is enhanced. Theasymptotic stability of the controller is proved by Lyapunov theory. Simulation andexperimental results show that the effectiveness of the control strategy on the SRV-02platform.The key to this design is the nonlinear parameter, has a strong promotional. In this design, theStribeck model was used to describe friction, but this design method has a strong applicationwhich can be extended to the dynamic model and other adaptive systems.