Research On The Motion Control Of Humanoid Leg Joints Actuated By Pneumatic Artificial Muscles

Pneumatic products are increasingly applied to a wide range scope of social life, industrial production due to its unique properties. As a new type of pneumatic actuators, pneumatic artificial muscle has the good performance similar to the nature muscle and is causing more and more attention. It is foreseeable that extremely good prospects in the therapy, bionics, and some other fields.This work took the bionic application of pneumatic muscles as the research target. The mechanical structure of humanoid leg joints actuated by pneumatic artificial muscles was designed. Based on the experimental results and theoretical model, the mathematical models of pneumatic artificial muscles, high-speed solenoid on-off valves were established. The modeling and analysis of single degree of freedom (DOF) and two-DOF pneumatic muscles joint were operated respectively. Several robust control methods including sliding mode control (SMC) and adaptive robust control (ARC) algorithm were applied to achieve high precision tracking performance. The paper is composed of seven chapters; main contents of each chapter are described as followed.The chapterâ… describe the issues related to the research background. The working principle and development of pneumatic muscles was introduced. Reviewed the research achievements and developing trends in the relevant fields at home and abroad. The origin and significance of the work is given and then the main work and research content are noted.Chapterâ…¡established the hardware system for research. Based on human physiological structure of the lower limb, designed a multi-DOF humanoid leg joints structure actuated by pneumatic muscles. To satisfy the needs of mobile robot application, Structural parameters has been optimized. Determine the corresponding components to meet the requirements of the control system. Two types of controller design platforms were established. And a preliminary experiment was conducted testing the feasibility of the whole system. Chapterâ…¢modeled the multi-DOF humanoid leg joint system. Taking the Germany company Festo's fluidic muscles as specific subjects, the related physical properties of a certain type pneumatic muscles were tested including the elastic modulus of its external rubber layer, various mechanical properties under the conditions of isometric, isotonic and isobaric. The large gap between the static theoretical model and experimental results was analysis, and from the controller design perspective, modified the model parameters. Then a higher precision static model was established. The high-speed PWM switching valves' flow characteristics was tested. Set up a single DOF and two-DOF pneumatic muscle joint's mathematical model based on the thermodynamics equations, as well as joint combined kinetic model. The static and dynamic experimental results demonstrated that the mathematical model is close to the experimental results. The dynamic performance of pneumatic muscle joint and some related factors were analysis by simulation.Considering the strongly nonlinear of pneumatic muscles joint and difficult to build a precise mathematical model, Chapterâ…£designed a sliding variable structure robust controller sliding mode controller on the basis of analysis of the characteristics and some simplifications. Simulation and experimental results show that the variable structure controller has good robustness and stability meeting the uncertain parameters and external disturbance. Steady-state error is less than 0.005 rad, low-frequency tracking error of single DOF joint is less than 0.04 rad, high-frequency tracking error is less than 0.06 rad. The two-DOF joint's low-frequency tracking error is less than 0.025 rad, high-frequency tracking error is less than 0.06 rad.In order to further improve control precision, Chapterâ…¤applies a robust adaptive control (ARC) strategy based on robust control design which integrated the adaptive control and robust control strategy. It guaranteed the stability of the closed-loop system and at the same time further reduces the tracking error and increase system performance by adjusting the model parameters adaptively. The experimental results show that the steady-state error is less than 0.003rad, single DOF joint's low-frequency tracking error less than 0.01 rad, high-frequency tracking error is less than 0.05rad. Low-frequency error of two DOF joint is less than 0.015rad, high-frequency tracking error is less than 0.03 rad.In order to reduce the cost of practical application, Chapterâ…¥adopt the saturation robust adaptive controller (SARC) using the position feedback only. It is still on the basis of robust control, but by combining parameters adaptive laws improve the closed-loop control accuracy. As the pressure feedback is not need, thus the utility of the control system is greatly enhanced. The experimental results show that for the single DOF joint, the low-frequency continuous tracking error is less than 0.01 rad, high-frequency tracking error is less than 0.03 rad, and for two-DOF joint low trajectory tracking error is less than 0.025 rad, high-frequency tracking error is less than 0.03 rad.Chapterâ…¦made the summation, some results and conclusions of the study were presented, and future research work were given.