老年人下肢运动能力建模与人-机运动交互关键技术研究

In the current Chinese society, the problem of aging is becoming more and more serious. The maintenance of the lower limb physical ability of the elderly is the premise for a high quality life. This project carries out the objective evaluation method for the elderly's physical ability and control algorithms for human-machine steady motion. The main contents are as follows: Firstly, on the base of activity information and video information, the objective evaluation model of the elderly's lower limb physical ability is established, taking off language description and guidance of experienced physician. Secondly, sectional dynamic models of the swing and support phase, containing information of the auxiliary walking equipment and human upper limbs, are proposed. The model can provide a basement for the steady control of the human-machine system. Thirdly, by using of laser signal and human activity information, the desired output signal for the auxiliary walking equipment is acquired accurately and rapidly. Lastly, to improve the steady of the auxiliary walking equipment users during walking in different speeds and different road conditions, the estimation of the interaction force between man and machine system and the constrained sliding mode control algorithm are designed. . The research of this project will not only enhance theoretical value in the elderly's daily travel ability, but also provide a favorable theoretical basis and technical support for the development of auxiliary walking equipments and exoskeleton devices.

当前中国社会，老龄化问题日趋严重，老年人下肢运动功能的保持是其拥有高品质老年生活的前提保障。本项目针对老年人这一特殊群体，开展下肢运动能力客观评价方法和穿戴智能辅助行走设备时人-机平稳运动控制方面的研究。研究内容主要包括：1、根据人体运动物理信息和视频信息，建立人体运动能力客观评价模型，脱离开语言描述和有经验医师指导。2、建立包含助行设备和人体上肢信息的人-机系统摆动期和支撑期分段动力学模型，为平稳控制方法提供保障。3、利用激光信号和人体运动物理信号，准确、快速获得控制智能辅助行走设备所需的期望输出信号。4、通过对人-机系统相互作用力的估计和约束型滑模控制器的设计，帮助老年人在不同步速、不同路况下平稳行走。. 本项目不仅为老年人日常出行能力的提升提供了必要的理论基础，同时也为助行设备和外骨骼装置研发提供了有利的理论依据和技术支持。

人口老龄化日趋严重是我国社会经济发展中的一个重要问题，老龄化人群肌力衰退，活动能力下降，下肢助行外骨骼能够有效帮助老年人恢复运动能力，提高生活品质，已成为解决老龄化困境的一种重要技术手段，具有极高的理论研究意义和社会应用价值。.本项目针对人体运动能力难以客观准确判别的问题，提出了将Xception-LSTM与核主成分分析结合、改进条件深度置信网络与长短时记忆网络相结合的两种下肢运动能力评价方法。采用激光雷达信息，设计了FWA-XGBoost算法实现对人体运动意图的识别。建立了包含助行外骨骼和人体下肢信息的人-机系统摆动期和支撑期分段动力学模型，为下肢外骨骼的平稳跟踪控制提供保障。针对外骨骼动力学模型不准确和人-机交互力的问题，提出了一种自适应滑模控制策略。通过对下肢助行外骨骼模型进行近似离散动态线性化，基于偏格式动态线性化中的伪梯度设计了外骨骼平稳行走离散终端滑模控制器。项目中对所提控制方法均进行了稳定性分析。.本项目为人体运动能力准确评价和肢体障碍者或老年人助力出行等难点问题的解决提供了有效途径和新的思路。在项目执行期间，发表了高水平学术论文21篇；申请国家发明专利11项，已授权4项。本项目产生的具有自主知识产权的成果，可以广泛应用到机器人、可穿戴外骨骼、智能康复系统、智能康养等多个领域。