Following assistance control of walking assist exoskeleton based on oscillator
                                          YANG Wei1,2, YANG Canjun1,2, MA Zhangyi1, WANG Hansong1, XU Linghui1
（1.College of Mechanical Engineering, Zhejiang University, Hangzhou 310027, China;2.Ningbo Research Institute, Zhejiang University, Ningbo 315100, China）


Abstract: Aiming at the problem of humanexoskeleton uncoordinated interaction, study on human walking intention recognition methods and hip exoskeleton joint compliant assist torque control were carried out. Through gait characteristics analysis of normal human walking, a phase oscillator based walking intention recognition method was designed. With simulations and experiments of test platform, phase oscillator based exoskeleton following assistance control theory and its performace were verified. Meanwhile, optimizational algorithms were verified on test platform to realize exoskeleton initial state selfcalibration and static state antishake. Finally, 10 meters walking tests were conducted on 5 volunteers. The results indicate that with walking assist exoskeleton, various volunteers average walking speed is increased to 20.16% compared with free walking. In addition, each volunteer gave positive feedback on walking experience without sudden shake of assist force. Phase oscillator based walking assist exoskeleton following assistance control can not only adapt to wearer’s walking stride and speed, but also provide effective and compliant hip joint assist torque, realize smooth and normal walking.
Key words: walking assist exoskeleton; oscillator; motion intention; following assistance