Abstract：With two antagonistic pneumatic muscle joints in series, a 2-DOF (degree of freedom) leg mechanism is designed, and the Lagrange dynamics model of the single leg is developed based on the pneumatic muscle model with three factors. The system state space is established, and the sliding mode control (SMC) law based on the upper bound of the interference is designed. The simulation and experimental platforms are built. Simulations and experiments of position tracking for hip and knee joints are completed with PID (proportional-integral-differential) control and SMC. The experimental results show that compared with PID control, the position errors of hip and knee joints are reduced by 26.2% and 25.1% respectively under SMC. So, the position tracking accuracy of SMC is better than PID control. The experiment is carried out on the self-developed platform of quadruped robot, which can realize the trot gait under SMC control.
 Tondu B. Modelling of the McKibben artificial muscle:A review[J]. Journal of Intelligent Material Systems and Structures, 2012, 23(3):225-253.
 Chang M K, Liou J J, Chen M L. T-S fuzzy model-based tracking control of a one-dimensional manipulator actuated by pneumatic artificial muscles[J]. Control Engineering Practice, 2011, 19(12):1442-1449.
 Chang M K. An adaptive self-organizing fuzzy sliding mode controller for a 2-DOF rehabilitation robot actuated by pneumatic muscle actuators[J]. Control Engineering Practice, 2010, 18(1):13-22.
 Hashimoto Y, Nakanishi Y, Saga N, et al. Development of gait assistive device using pneumatic artificial muscle[C]//Joint 8th International Conference on Soft Computing and Intelligent Systems (SCIS) and 17th International Symposium on Advanced Intelligent Systems. Piscataway, USA:IEEE, 2016:710-713.
 Wu Q C, Wang X S, Du F P, et al. Fuzzy sliding mode control of an upper limb exoskeleton for robot-assisted rehabilitation[C]//IEEE International Symposium on Medical Measurements and Applications. Piscataway, USA:IEEE, 2015:451-456.
 Chang M K. Tracking control of a leg rehabilitation machine driven by pneumatic artificial muscles using composite fuzzy theory[J]. The Scientific World Journal, 2014(2014):No.464276.
 Zhao L, Ge L L, Yang Y F. Active disturbance rejection trajectory tracking control for two-joint driven by pneumatic artificial muscles[C]//2015 International Conference on Fluid Power and Mechatronics. Piscataway, USA:IEEE, 2015:972-977.
 Eker ?. Second-order sliding mode control with experimental application[J]. ISA Transactions, 2010, 49(3):394-405.
 王杨,张强,肖晓晖.基于鲁棒建模的气动人工肌肉驱动仿生关节的轨迹跟踪控制[J].机器人,2016,38(2):249-256.Wang Y, Zhang Q, Xiao X H. Trajectory tracking control of the bionic joint actuated by pneumatic artificial muscle based on robust modeling[J]. Robot, 2016, 38(2):249-256.
 陶国良,谢建蔚,周洪.气动人工肌肉的发展趋势与研究现状[J].机械工程学报,2009,45(10):75-83.Tao G L, Xie J W, Zhou H. Research achievements and development trends of pneumatic artificial muscles[J]. Journal of Mechanical Engineering, 2009, 45(10):75-83.
 鲍春雷,王斌锐,金英连,等.气动肌肉关节的无模型自适应CMAC迟滞补偿控制[J].机器人,2015,37(3):298-303,309.Bao C L, Wang B R, Jin Y L, et al. Model free adaptive CMAC hysteresis compensation control of the pneumatic muscle joint[J]. Robot, 2015, 37(3):298-303,309.
 Lilly J H, Quesada P M. A two-input sliding-mode controller for a planar arm actuated by four pneumatic muscle groups[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2004, 12(3):349-359.
 Serres J L, Reynolds D B, Phillips C A, et al. Characterisation of a phenomenological model for commercial pneumatic muscle actuators[J]. Computer Methods in Biomechanics and Biomedical Engineering, 2009, 12(4):423-430.
 蔡自兴,谢斌.机器人学[M].3版.北京:清华大学出版社,2015:79-83.Cai Z X, Xie B. Robotics[M]. 3rd ed. Beijing:Tsinghua University Press, 2015:79-83.
 马明,王建中,盛沙.基于虚拟样机技术的4足机器人机构设计[J].机械设计,2012,29(3):25-28.Ma M, Wang J Z, Sheng S. Mechanism design of quadruped robot based on virtual prototyping technology[J]. Machine Design, 2012, 29(3):25-28.