Posture Control Of Parallel Platform Driven By Piston And Pneumatic Muscles

Pneumatic parallel platforms have many advantages such as high power-mass ratio, simple structure, low price, cleanness and easy maintenance, so they are widely applied in bionic and industrial robots, medical and rehabilitation equipment, shock absorbing and other industrial fields. As pneumatic servo systems have shortcomings such as low stiffness and nonlinear characteristics, which bring difficulty in high-precision motion control, the posture trajectory tracking control of pneumatic parallel platforms has been an important research direction. In this dissertation a three-DOF pneumatic parallel platform which is driven by an air cylinder and three pneumatic muscles is taken as the research object. To accomplish the high-precision posture trajectory tracking control, a combination of theoretical, emulational and experimental research methods is employed. The model of the pneumatic parallel platform is analysed firstly, then the control strategy of trajectory tracking control of the pneumatic parallel platform and the synchronous control strategy of multiple pneumatic parallel platforms are further studied.To achieve the high-precision motion control of the pneumatic parallel platform, the force character of pneumatic muscles is modeled and analysed by experiments firstly. An adaptive robust controller which controls stiffness and posture of the parallel platform separately is designed through back-stepping method, and is applied in simulation experiments. On this basis, a numerical posture solver is designed to calculate the posture of the moving plate from lengths of pneumatic muscles, due to the fact that the parallel platform has no analytical solution of the forward kinematics. Then an integral adaptive robust controller is designed in working space to accomplish the posture trajectory tracking control of the parallel platform. Meanwhile, an integral cross-coupling adaptive robust controller which combines cross-coupling control strategy and adaptive robust control theory is designed in joint space. Experiments show that these two controllers can both achieve a high precision of posture trajectory tracking control, and have the ability of parameters estimation and control robustness.In addition, to accomplish the synchronous control of multiple pneumatic parallel platforms in the same structure, a splited cross-coupling adaptive robust control strategy is designed to reduce the synchronous controlling error and improve flexibility of the array of parallel platforms.This doctoral dissertation consists of seven chapters as follows.In chapter 1, the researching background related to pneumatic servo parallel platform is reviewed. Research significance, challenges and main contents of this dissertation are illustrated.In chapter 2, the force character of pneumatic muscles is analysed through experiments. A modified formula is proposed to improve modeling accuracy of pneumatic muscles. Metal wire is filled in the robber bladder of pneumatic muscles to reduce slow time-varying characteristics. The pneumatic servo system of the parallel platform including mechanical structure, pneumatic system and servo controller is designed on this basis. Structural parameters and types of components are chosen properly to set up the experimental device.In chapter 3, dynamic character of the pneumatic parallel platform and modeling of servo controlling valves and actuators are analysed in MATLAB/Simulink. An adaptive robust controller which controls stiffness and posture of the parallel platform separately is designed based on the strong coupling and highly nonlinear character, and is applied in simulation experiments. Simulation results show that the controller can achieve a high level of posture tracking controlling precision, and has the ability of parameters estimation and control robustness as well.In chapter 4, as the parallel platform has no analytical solution of forward kinematics, a numertical posture solver based on the quasi-newton method is designed and verified through simulation experiments. An integral direct-indirect adaptive robust controller is designed due to slow time-varying character of pneumatic muscles to accomplish the posture trajectory tracking control. The least square method is applied in an indirect adaptive observer of the controller to identify the force parameters of pneumatic muscles in real time. The stability of the controller is proved theoretically. The integral adaptive robust controller and the adaptive observer is verified through a series of experiments. Compared with traditional adaptive robust controllers, the controller with integral feedback term has higher posture tracking control accuracy. The universal adequacy and robustness of the controller are also verified through experiments.In chapter 5, to avoid occupying too much computing resource by the numertical posture solver, an integral cross-coupling direct-indirect adaptive robust controller is designed in joint space. Coupling errors are employed to take place of the posture errors in joint space according to the cross-coupling control strategy. The least square method is applied in the indirect adaptive observer of the controller. The stability of the proposed controller is proved theoretically. A series of controllers are designed to testify the function of integral feedback term and cross-coupling strategy. Experimental results show that the integral cross-coupling adaptive robust controller can achieve high presicion and robustness in posture trajectory tracking control of the parallel platform. The integral feedback term and the cross-coupling strategy can improve trajectory tracking performance of the parallel platform significantly.In chapter 6, the cross-coupling control strategy is employed to reduce the synchronous control error of multiple parallel platforms. A two-platform cross-coupling adaptive robust controller is designed firstly. The stability of the controller is proved theoretically. To control an array of parallel platforms, a multiple cross-coupling adaptive robust controller is designed based on separation principle. The validity of the proposed controllers and cross-coupling strategy is verified through simulation experiments. Simulation results show that the cross-coupling control strategy can significantly reduce the synchronous posture tracking error of multiple parallel platforms. The cross-coupling adaptive robust controller can converge the posture error of single parallel platform and the synchronous error between multiple parallel platforms simultaneously. The cross-coupling controllers also have the ability of parameters estimation and control robustness.In chapter 7, the conclusions of the dissertation are illustrated. The main contributions and innovative points are summarized, and the recommendations of future study are proposed.