| With the rapid development of mechanization and automation,more and more attention and research have been paid to the pneumatic system.Because of its low cost,clean and safety,simple structure,convenient maintenance,the pneumatic system has been widely applied in medical,robot,food packaging,mechanical operation and automobile manufacturing.However,due to the disadvantages of the pneumatic system,such as nonlinearity of the servo valve,friction and air compressibility,it is difficult to get satisfactory positioning accuracy.Therefore,how to improve the positioning accuracy of the pneumatic system is one of the important research directions for the current pneumatic technology.This paper will take the rod cylinder servo system as the research object,and use the active disturbance rejection control technology to study the positioning accuracy and the response speed.Firstly,through the study and analysis for the structure and working principle of the rod cylinder servo system,the corresponding experimental platform is set up.According to Newton’s second law theory and so on,the mathematical model of the rod cylinder servo system is obtained by reasonable assumption and simplification in the process of derivation.However,due to the uncertain problems of the system such as complex nonlinearity,unmodeled dynamics and external environmental disturbances,it is difficult to establish a complete mathematical model.Secondly,in order to solve the various uncertainties of the rod cylinder servo system,a linear active disturbance rejection control method optimized by least squares support vector machine is applied to the positioning research of the system.The linear extended state observer and the least squares support vector machine are used to estimate the various uncertainties of the system in real time.The linear state error feedback controller can ensure good performances of the closed-loop system by the compensation of the estimation value.In addition,it is proved by Lyapunov theory that the estimation error of linear extended state observer and the positioning error of closed-loop system are bounded,and it decreases with the increase of their eigenvalues.The simulation results verify thatthe linear active disturbance rejection control method optimized by least squares support vector machine can improve the positioning effect of rod cylinder servo system well.Thirdly,due to the above control method does not consider the finite time problem,a finite time positioning control method based on an extended state observer is proposed for the rod cylinder servo system.The extended state observer with finite time convergence is constructed to estimate the uncertainties of the system.The nonsingular fast terminal sliding mode controller is designed to ensure good performances of the closed-loop system.In addition,it is proved by the analysis of Lyapunov theory that the sufficient small estimation error and the stabilization of the closed-loop system are obtained in finite time.The effectiveness of the proposed control method is verified by comparative experimental researches with the linear active disturbance rejection control method.Finally,the generalized extended state observer is proposed to deal with the positioning problem of the rod cylinder servo system with varying load.Compared with the traditional extended state observer,the generalized extended state observer can not only estimate various uncertainties and varying load of the system,but also estimate their differential.Therefore,the observer can estimate the high-order uncertainty of the system more accurately and compensate the estimation value to the integral sliding mode controller to ensure the positioning effects of the closed-loop system.At the same time,the convergence of the generalized extended state observer and the stabilizability of the closed-loop system are analyzed in finite time by Lyapunov theory.The experimental results show that the proposed control method is reasonable and effective in the application of the rod cylinder servo system with varying load. |
