多源干扰外骨骼机器人系统的复合分层抗干扰控制研究

Exoskeleton robot is a kind of robot which is worn in human body to help human bear load and assist human in walking. It has great value of application in the military, disaster rescue, medical care and other fields. In practical application, however, disturbances, including parameter perturbation of system, unmodeled dynamics and external disturbance, etc., are the major factors which degrade the control performance of exoskeleton robot. In this project, the main work is to improve the control performance of exoskeleton robots under the multiple sources of disturbances. After extensive analyses of the multiple sources of disturbances in exoskeleton robots, the model of disturbance of exoskeleton robot has been established. For parameter perturbation of system, the technology of parameter identification has been employed to estimate the value of disturbances, and the estimated value is used to compensate specific disturbances online. Aiming at unmodeled dynamics and external disturbance, according to different type of disturbances, the technology of disturbance observer has been used to estimate the disturbances of system. Considering variations of working condition, requirements of fast dynamic performance and integrated optimization, different feedback control methods including adaptive control, finite time control, model predictive control, are developed respectively. Last, for multiple sources of disturbances in exoskeleton robot system, a hierarchical anti-disturbance control scheme has been proposed to suppress the multiple sources of disturbances. Research achievements of this project will effectively improve the dynamic performance, tracking precision, disturbance rejection ability of exoskeleton robot closed loop system. In addition, this study provides a basic theory and technology supporting materials for the further research works in the aspect of exoskeleton robot control system.

外骨骼机器人是一类能够增强和改善人体运动机能的穿戴式设备，在军事、医疗等领域有着重要的应用价值。在实际应用中，扰动是影响外骨骼机器人控制性能的主要因素，包括: 参数摄动，未建模动态以及外部扰动等。本项目以提升多源干扰情况下外骨骼机器人的控制性能为目标展开研究。在深入分析外骨骼机器人系统工作环境中多种干扰特征的基础上，建立系统的干扰模型。针对外骨骼机器人系统参数摄动，设计相应的模型参数辨识方法。针对外骨骼机器人系统的未建模动态和外部扰动，设计扰动观测方法。针对工况自适应、快速动态性能、综合优化性能需求，分别设计相应的自适应控制、有限时间收敛控制、模型预测控制等反馈控制策略。最后，针对多源扰动外骨骼机器人系统，采用复合分层抗干扰控制方法，实现多源扰动的主动抑制。项目研究成果将有效提升外骨骼机器人闭环系统的动态、跟踪精度、抗干扰等性能，为进一步研制高性能的外骨骼机器人系统提供理论依据和技术支撑。

本项目以外骨骼机器人系统为研究对象，以提高复杂工况下外骨骼机器人系统抗干扰性能为目的，采用理论分析、仿真与实验验证相结合的研究方法对外骨骼机器人系统的多源干扰进行建模与主动抑制。首先，采用机理分析和实验测试相结合的方法对外骨骼机器人系统中的多源干扰进行具体的分析和建模。针对系统中存在的参数摄动问题，设计相应的模型参数辨识方案；针对系统中存在的未建模动态以及外部扰动问题，设计高精度的扰动观测器。针对工况自适应、快速动态性能、综合优化性能需求，分别设计相应的自适应控制、有限时间收敛控制、模型预测控制等反馈控制策略。最后，结合干扰前馈补偿与反馈控制技术，设计复合分层结构的抗干扰控制策略，对外骨骼机器人系统的多源扰动进行抑制。项目研究可为外骨骼机器人控制系统的设计提供重要的基础理论支持。