本文選題：迭代學(xué)習(xí)控制 + 模型預(yù)測控制��； 參考：《東北大學(xué)》2014年碩士論文

【摘要】：腦卒中(腦中風(fēng))是嚴重危害人類健康和生命安全的常見的難治性疾病,存在著明顯的三高(發(fā)病率高、致殘率高、死亡率高)現(xiàn)象。腦中風(fēng)幸存患者多數(shù)有不同程度喪失勞動力和生活不能自理的情況,致殘率高達75%,給患者及其家庭帶來了沉重的負擔(dān)。腦卒中引發(fā)的肢體運動障礙患者經(jīng)過正規(guī)的康復(fù)訓(xùn)練可以明顯減少或減輕癱瘓的后遺癥,因此有效的腦卒中患者的康復(fù)訓(xùn)練手段是亟需研究的課題。功能性電刺激技術(shù)(Functional Electrical Stimulation, FE S)是近年來迅速發(fā)展起來的一門嶄新的醫(yī)療技術(shù),其在腦卒中患者的肌肉鍛煉以及神經(jīng)通路的功能恢復(fù)方面效果顯著,引起了眾多醫(yī)療科研人員的關(guān)注。本文針對腦卒中患者上肢康復(fù)問題開展研究,將鏡像康復(fù)方法與功能電刺激相結(jié)合,深入全面地分析了上肢鏡像康復(fù)運動控制方法。本文的主要內(nèi)容包括以下幾個方面：首先,為最大程度降低FES引起的肌肉疲勞,本文對功能性電刺激不同參數(shù)調(diào)節(jié)對肌肉疲勞的影響進行了比較研究。功能性電刺激引起的肌肉疲勞已經(jīng)成為阻礙功能性電刺激應(yīng)用推廣的主要限制因素,而降低功能性電刺激肌肉疲勞能夠使得訓(xùn)練康復(fù)更加高效。為定量評估刺激脈寬、頻率和幅值三種參數(shù)變化情況下引起的肌肉疲勞狀況,本文創(chuàng)新性地提出基于表面肌電信號(sEMG)的肌肉疲勞強度判斷準(zhǔn)則,最后通過對實際人體FES刺激數(shù)據(jù)的分析總結(jié)得出了脈寬調(diào)節(jié)對肌肉疲勞影響較小的結(jié)論。其次,針對腦卒中患者肌肉力量不足,容易疲勞的特點,開展了基于脈寬調(diào)節(jié)功能性電刺激的離散動作鏡像康復(fù)方法研究。通過KINECT捕捉患者健康側(cè)的運動,將運動鏡像到患側(cè)肢體,結(jié)合FES使受控制的患側(cè)肢體與健側(cè)肢體以相同速度完成動作,實現(xiàn)離散鏡像康復(fù)訓(xùn)練。為保證患側(cè)肢體動作的準(zhǔn)確性,提出了基于迭代學(xué)習(xí)控制(Iterative Learning Control, ILC)與PD反饋相結(jié)合的FES控制方法。通過對健康被試在不同速度(慢速、中速和快速)下的實驗表明所提的方法對于患側(cè)運動的跟蹤取得了良好的效果。最后,針對具有一定肌肉力量和耐力的腦卒中患者,本文提出了基于功能電刺激的連續(xù)運動鏡像康復(fù)訓(xùn)練方法,利用功能性電刺激控制患側(cè)肢體實時跟蹤健側(cè)肢體的連續(xù)運動。本文針對人體肌肉骨骼特點,提出了基于模型預(yù)測控制(Model PredictiveControl, MPC)的患側(cè)肢體跟蹤控制方法。仿真和實驗結(jié)果表明,模型預(yù)測控制能對患側(cè)肢體運動實現(xiàn)良好的控制效果。綜上,通過結(jié)合功能性電刺激以及鏡像康復(fù)兩種腦卒中康復(fù)手段,提出了基于功能性電刺激的離散動作鏡像康復(fù)及連續(xù)動作鏡像康復(fù)兩種訓(xùn)練方法,仿真以及實驗結(jié)果表明方法簡單且有效,對于發(fā)展新型腦卒中康復(fù)技術(shù),提升腦卒中康復(fù)效果起到了積極的促進作用。
[Abstract]:Cerebral apoplexy (stroke) is a common refractory disease that seriously endangers human health and life safety. Most of the survivors of stroke are unable to take care of themselves to some extent, and the disability rate is as high as 75, which brings a heavy burden to the patients and their families. Regular rehabilitation training can significantly reduce or alleviate the sequelae of paralysis in patients with motor disorders caused by stroke. Therefore, effective rehabilitation training for stroke patients is a subject that needs to be studied. Functional Electrical stimulation (FE S) is a new medical technique which has been developed rapidly in recent years. It has a remarkable effect on muscle exercise and functional recovery of nerve pathway in stroke patients. It has attracted the attention of many medical researchers. In this paper, the rehabilitation of upper limb in stroke patients was studied. The method of mirror rehabilitation and functional electrical stimulation was combined to analyze the motor control method of mirror rehabilitation of upper limb. The main contents of this paper include the following aspects: firstly, in order to minimize the FES induced muscle fatigue, the effects of different parameters of functional electrical stimulation on muscle fatigue were studied. Muscle fatigue caused by functional electrical stimulation has become the main limiting factor which hinders the application of functional electrical stimulation. Reducing functional electrical stimulation muscle fatigue can make training and rehabilitation more efficient. In order to quantitatively evaluate the muscle fatigue caused by the changes of pulse width, frequency and amplitude, a new criterion for judging muscle fatigue strength based on surface electromyography (SEMG) is proposed in this paper. Finally, the conclusion that pulse width regulation has little effect on muscle fatigue is concluded by analyzing the actual human FES stimulation data. Secondly, the discrete action mirror rehabilitation method based on pulse width regulation and functional electrical stimulation was carried out in view of the characteristics of insufficient muscle strength and easy fatigue in stroke patients. The movement of the healthy side of the patient was captured by KINECT, the movement was mirrored to the affected limb, and FES was combined with FES to complete the movement at the same speed as the healthy limb, and to realize the rehabilitation training of discrete mirror image. In order to ensure the accuracy of limb movement, an iterative learning control (ILC-based) and PD feedback based FES control method is proposed. Experiments on healthy subjects at different speeds (slow, medium and fast) show that the proposed method is effective in tracking the affected side motion. Finally, for stroke patients with certain muscle strength and endurance, a method of continuous motor mirror rehabilitation training based on functional electrical stimulation is proposed in this paper. Functional electrical stimulation is used to control the continuous movement of the affected limb in real-time tracking of the healthy limb. According to the characteristics of human musculoskeletal, a method of limb tracking control based on Model Predictive Control (MPC) is proposed. Simulation and experimental results show that the model predictive control can achieve a good control effect on the motion of the affected limb. On the basis of functional electrical stimulation and mirror rehabilitation, two training methods, discrete action mirror rehabilitation and continuous action mirror rehabilitation, were put forward. The simulation and experimental results show that the method is simple and effective, which plays a positive role in the development of new stroke rehabilitation technology and the improvement of stroke rehabilitation effect.
【學(xué)位授予單位】：東北大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：R743.3

【參考文獻】

相關(guān)期刊論文 前4條

1 王維;馬躍文;楊巍;;鏡像療法對腦卒中偏癱患者上肢功能和手功能的影響[J];大連醫(yī)科大學(xué)學(xué)報;2013年06期

2 林海波;梅為林;張毅;羅元;;基于Kinect骨骼信息的機械臂體感交互系統(tǒng)的設(shè)計與實現(xiàn)[J];計算機應(yīng)用與軟件;2013年02期

3 王建輝;張傳鑫;白冰;劉一楠;;上肢康復(fù)機器人實驗平臺肌電信號采集系統(tǒng)的設(shè)計[J];電子世界;2012年11期

4 鄒智;張英;王珊珊;廖維靖;;鏡像治療結(jié)合任務(wù)導(dǎo)向性訓(xùn)練對腦卒中患者上肢功能的影響[J];中華物理醫(yī)學(xué)與康復(fù)雜志;2011年09期

相關(guān)博士學(xué)位論文 前2條

1 成娟;基于表面肌電和加速度信號融合的動作識別和人體行為分析研究[D];中國科學(xué)技術(shù)大學(xué);2013年

2 馬長波;基于電機輔助功能性電刺激腳踏車康復(fù)訓(xùn)練系統(tǒng)的研究[D];哈爾濱工業(yè)大學(xué);2010年

相關(guān)碩士學(xué)位論文 前1條

1 楊瑞;肌電信號采集與分析系統(tǒng)的研制[D];華中科技大學(xué);2009年

，

本文編號：2010723