可穿戴式下肢助行机器人摔倒预测及其稳定恢复控制研究

The ability of stability recovering from falling trends play very important role on the security of the wearer in the wearable lower limb robot system of walking assist for persons with weak balancing capability. However the existing stable control methods of the wearable robot system mainly focus on recovering from very small disturbance, and seldom deal with the case of stronger disturbance.In order to realize the stability reovering from stronger disturcance,falling prediction method and stability recovering control for strong disturbance will be studied deeply in this project. Firstly, the quantitative description method of the stability for each stage during the falling will be proposed acoording to the analysing result of the dynamic stable of the man-machine system. Then the relation between the falling behavior and the signals from sensors will be analyzed based on principles of statistics, after that the characteristic parameters will be extracted from the signals. Modeling method of falling prediction model based on time series as well as the parameter optimization method for the prediction model will be researched deeply. Lastly, the essence of the balance strategies human adopted will be studied, and then man-machine combination stability control strategies will be proposed, in the initial security stage of the falling ,the stability will be recovered by the wearer, the robot was controlled passively following to the wearer, during the other stages of falling, the robot control actively to finish stability recover, and the control realization for each stability recovering strategy will be done in this project. The research results of the project can not only realize the falling predicting and the stability recovering controlling, but also provide a strong theoretical basis and methodological guidance for the application of the wearable robot in the walking helping for aged persons.

可穿戴式下肢助行机器人的稳定恢复性能，对平衡能力较弱的穿戴者的安全性至关重要。本项目拟针对大干扰下的稳定恢复，开展摔倒预测方法、摔倒过程中稳定恢复控制技术研究。结合系统动力学稳定原理研究，明确摔倒过程各阶段稳定性的内涵及定量描述方法；研究摔倒过程特征参数表征方法，开展摔倒模型影响要素分析,研究建立基于时间序列建立摔倒预测模型和模型参数优化方法；深入研究人体摔倒过程中稳定平衡策略的科学根源，提出摔倒初始安全阶段穿戴者主动调整、机器人人被动跟随，摔倒其他阶段机器人主动控制的人机结合的稳定恢复控制策略，通过运动学及动力学分析，研究稳定恢复控制策略的实现方法。通过项目研究，可实现摔倒预测及大干扰下的稳定恢复。研究成果可深化对摔倒过程稳性的认知,丰富稳定性控制方法,可为穿戴式外骨骼下肢助行机器人助老应用，提供更有利的理论依据和方法指导。

可穿戴式下肢助行机器人的稳定恢复性能，对平衡能力较弱的穿戴者的安全性至关重要。本项目普通的助力机器人控制算法没法满足大干扰下助力机器人系统的稳定性恢复问题，针对大干扰下的稳定恢复，开展了摔倒预测方法、摔倒过程中稳定恢复控制技术的研究。结合系统动力学稳定原理研究，明确摔倒过程各阶段稳定性的内涵及定量描述方法；根据动力学，对摔倒阶段进行了划分；研究并建立了基于时间序列的摔倒预测模型和模型参数优化方法；深入研究了人体摔倒过程中稳定平衡策略的科学根源，提出了摔倒初始安全阶段穿戴者主动调整、机器人人被动跟随，摔倒其他阶段机器人主动控制的人机结合的稳定恢复控制策略，通过运动学及动力学分析，提出了稳定恢复控制策略并进行了仿真验证 。通过该项目研究，实现了摔倒的预测及大干扰下的稳定恢复控制。研究成果深化了对下肢外骨骼机器人系统摔倒过程稳性的认知,丰富稳定性控制方法,可为穿戴式外骨骼下肢助行机器人助老应用，提供更有利的理论依据和方法指导。