Nonlinear Friction Compensation Adaptive Control For Mechanical Servo Systems

The advancement of science and technology promotes the development of human economy.The desire of human beings to improve the standard of living is also stronger.Various equipments used in daily life are developing towards automation and intelligence.These achievements are inseparable from high-precision machinery.The problem of improvement of the accuracy of machinery and equipment has attracted much attention of the majority of scientific and technological workers.Non-linear friction is one of the key factors affecting the motion accuracy of mechanical servo systems,and it exists in almost all mechanical servo systems.Non-linear friction will increase the steady-state error of the system and thereby degrade the motion accuracy of the system.Therefore,how to solve the friction existing in the system has become an important issue in the field of machining and control.By using the position and velocity of the mechanical servo system,three state feedback control methods are proposed to compensate the nonlinear friction in the mechanical servo system.The three methods are simple and easy to implement in engineering.The specific work can be summarized as follows:First,based on the precise model parameters of the system,a state feedback robust proportional-derivative(RPD)plus servo system dynamics compensation(RPD+)control is proposed,where the robust control involving only on position error is used to achieve nonlinear friction compensation.Second,to overcome the weakness of RPD+ control that requires accurate known system dynamics with model parameters,parameter adaptation technique is introduced and a state feedback robust adaptive proportional-derivative(RAPD)control is developed.Third,to further improve the dynamic and static performances of the system,a class of nonlinear function is introduced and a state feedback robust adaptive nonlinear proportional-derivative(RANPD)control is proposed.The Lyapunov's direct method and Barbalat's lemma are used to prove the global asymptotic stability of the three closed-loop control systems.Numerical simulations with Matlab and experimental verifications on a Googol's single-degree-of-freedom mechanical servo system are performed to show the effectiveness and improved performances.The results show that the proposed control methods effectively compensate the non-linear friction existing in the mechanical servo system and the proposed RANPD control has better performances with faster convergence and smaller steady-state position tracking error.The modified continuous controls without chattering obtain the comparable results to the discontinuous controls.Meanwhile,the control method proposed in this thesis is compared with the widely-used sliding mode adaptive control method.The results show that the control method proposed in this thesis is simple in design process,easy to use,and has better trajectory tracking performance.