基于焦虑情绪与混杂控制的机器人辅助临床康复实验

发布时间：2018-08-03 07:50

【摘要】：现有基于患者主动参与的机器人辅助康复训练方法,大多以感知患者"运动"参与为主,而忽略了"心理"层次主动参与,且人机交互控制方法未能同时将机器人连续变量运动控制或医师离散事件决策控制这种混杂特性融于统一框架内,具有一定的局限。针对此问题,提出一种基于焦虑情绪识别与混杂控制的机器人辅助临床康复控制方法。该方法首先分析机器人辅助康复过程中不同强度焦虑情绪生理响应特征的显著性及差异性;其次,运用基于径向基核的支持向量机设计焦虑情绪分类器;再次,基于混杂控制理论设计与患者焦虑情绪强度相一致的人机交互控制器;最后,选取3例脑卒中患者,运用Barrett公司WAMTM机器人构建临床康复实验平台,验证所提方法有效性。
[Abstract]:Most of the existing robot-assisted rehabilitation training methods based on patient's active participation are mainly aware of the patient's "movement" participation, while ignoring the "psychological" level of active participation. The hybrid characteristics of robot continuous variable motion control or physician discrete event decision control can not be combined into the unified framework by man-machine interaction control method, which has some limitations. To solve this problem, a robotic assisted clinical rehabilitation control method based on anxiety recognition and hybrid control is proposed. This method firstly analyzes the significance and difference of physiological response characteristics of different intensity anxiety emotion in the process of robot assisted rehabilitation. Secondly, the support vector machine based on radial basis kernel is used to design anxiety emotion classifier. Thirdly, Based on hybrid control theory, a human-computer interaction controller is designed, which is consistent with the intensity of patients' anxiety. Finally, three stroke patients are selected to construct a clinical rehabilitation experimental platform using WAMTM robot of Barrett company to verify the effectiveness of the proposed method.
【作者单位】：南京邮电大学机器人信息感知与控制研究所;东南大学仪器科学与工程学院;
【基金】：国家自然科学基金(61673114,61305095) 江苏省重点研发计划(BE2015701) 江苏省自然科学基金(BK20141426、BK20170898) 江苏省“青蓝工程”项目(QL00516014)资助
【分类号】：R496;TP242

【相似文献】