基于卡尔曼滤波理论的KOA运动疗法规范性的研究

发布时间：2018-01-15 14:18

本文关键词：基于卡尔曼滤波理论的KOA运动疗法规范性的研究　出处：《福州大学》2014年硕士论文　论文类型：学位论文

【摘要】：膝骨性关节炎(Knee Osteoarthritis, KOA)是最常见的慢性、进展性的骨性关节疾病,此病进展缓慢,每个病人表现不完全相同,但最终都会表现为膝关节行走功能丧失。该疾病给患者心理和生理带来极大的创伤,使中老年人生活质量降低。因此安全、低成本地治疗KOA,具有广泛的社会需求和广阔的应用前景。运动疗法可以有效地改善关节功能,缓解关节疼痛,是一种效果良好且应用广泛的基础疗法。运动疗法的规范性是通过医生的经验指标来判断,要求患者在医院通过医生的指导进行运动疗法,在家难以确保患者运动的规范性,这就限制运动疗法的推广和应用。因此,本文为了增强患者在家进行运动疗法的训练效果,.采用基于滤波算法的人体下肢运动状态估计方法来进行研究,提出运动疗法规范性评判的新思路。本文研究内容主要包括以下几个方面：(1)加速度信号的采集。设计加速度采集系统采集加速度信号并将其传输到PC机进行处理与分析。(2)人体下肢运动模型的建立。引入加速度信号来描述人体下肢运动,对运动疗法中的直腿抬高(Straight Leg Raise, SLR)运动过程进行受力分析并建立状态空间模型,模型中的状态方程体现直腿抬高运动规律,其输出方程反映加速度信号和其状态变量间的关系。(3)滤波与状态估计。阐述卡尔曼滤波器、扩展卡尔曼滤波器(EKF)和基于Unscented变换的无迹卡尔曼滤波器(UKF)的原理、特点和实现步骤。对采集得到的直腿抬高运动数据进行预处理,并分别用EKF算法和UKF算法对人体下肢运动模型进行状态估计,最后并讨论这两种滤波器的滤波效果。结果表明,UKF滤波效果优于EKF。(4)规范性评判与监测。分析估计出的运动状态变量对规范性性能指标的影响程度,确定判断规则,进而进行规范性判断。最后设计基于Android操作系统的直腿抬高规范性监测软件,使得规范性结果可以传输到手机上并在其显示,实现了规范性的监测。实验结果表明,采用基于EKF和UKF的这两种人体下肢运动状态估计方法都可以客观定量地评判膝骨性关节炎的运动疗法规范性,并且通过直腿抬高规范性监测软件实现了运动疗法规范性的监测。
[Abstract]:Knee Osteoarthritis (KOA) is the most common chronic, progressive osteoarthropathy, which progresses slowly and is not identical in every patient. However, it will eventually be manifested as the loss of walking function of the knee joint. The disease brings great psychological and physiological trauma to the patients and reduces the quality of life of the middle-aged and elderly. Therefore, it is safe and low-cost to treat KOA. Exercise therapy can effectively improve joint function and relieve joint pain. Exercise therapy is a kind of basic therapy which has good effect and is widely used. The standard of exercise therapy is judged by the doctor's experience index and requires the patient to carry on the exercise therapy through the doctor's instruction in the hospital. It is difficult to ensure the normality of patients' exercise at home, which limits the popularization and application of exercise therapy. Therefore, this paper aims to enhance the training effect of exercise therapy at home. The method of estimation of human lower limb motion state based on filter algorithm is studied. A new idea for the normative evaluation of exercise therapy is put forward. The research contents of this paper mainly include the following aspects: 1). Acquisition of acceleration signal. Design acceleration acquisition system to collect acceleration signal and transmit it to PC for processing and analysis. The establishment of human lower extremity motion model. The acceleration signal is introduced to describe the human lower limb motion. The state space model of straight Leg raised straight Leg (SLR) in exercise therapy was established. The equation of state in the model reflects the law of straight leg elevation, and the output equation reflects the relationship between acceleration signal and its state variable. The principle of extended Kalman filter (EKF) and unscented Kalman filter (UKF) based on Unscented transform. Characteristics and implementation steps. The data of straight leg elevation are preprocessed, and the state of human lower limb motion model is estimated by EKF algorithm and UKF algorithm respectively. Finally, the filtering effects of these two filters are discussed. The effect of UKF filter is better than that of EKF.4) normative evaluation and monitoring. The influence of the estimated motion state variables on the normative performance index is analyzed and the judgment rules are determined. Finally, the standard monitoring software based on the Android operating system is designed to make the normative results can be transmitted to the mobile phone and displayed on the mobile phone. The results show that the two methods of motion state estimation based on EKF and UKF can objectively and quantitatively judge the standard of exercise therapy for knee osteoarthritis. And through straight-leg elevation normative monitoring software to achieve the normative monitoring of exercise therapy.
【学位授予单位】：福州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R684.3;TN911.7

【相似文献】