顧春榮

摘要：有別于基于圖像的計(jì)算機(jī)視覺技術(shù)和測(cè)量人體部位運(yùn)動(dòng)物理量的技術(shù)，肌電信號(hào)直接反映了人體的動(dòng)作意圖，從中提取的相關(guān)信息可轉(zhuǎn)化為控制輪椅的信號(hào)。該文敘述了肌電信號(hào)在操控智能輪椅中的主要工作，重點(diǎn)和發(fā)展方向。

關(guān)鍵詞：智能輪椅；人機(jī)接口；肌電信號(hào)；特征提取；分類識(shí)別

中圖分類號(hào)：TP18文獻(xiàn)標(biāo)識(shí)碼：A文章編號(hào)：1009-3044(2012)22-5442-04

Applications of Electromyographic Signal in Intelligent Wheelchair

GU Chun-rong

(College of Electronics and Information Engineering,Tongji university, Shanghai 201804, China)

Abstract: Differing from the image-based computer vision technique and the technique of measuring the physical caused by human parts motion, the electromyographic signal reflects the behavior intention of humans body directly. The relational messages extracted from the electromyographic signal can be converted into signals for controlling the wheelchair. This paper recites the primary works, emphasis and direction in steering the wheelchair using electromyographic signal.

Key words: intelligent wheelchair; Man-machine interface; electromyographic signal; feature extraction; classification and identification

智能輪椅作為醫(yī)療護(hù)理領(lǐng)域的服務(wù)機(jī)器人主要用來輔助老年人和殘疾人的日常生活和工作，是對(duì)他們?nèi)趸臋C(jī)體功能的一種補(bǔ)償。智能輪椅是將智能機(jī)器人技術(shù)應(yīng)用于電動(dòng)輪椅，融合機(jī)器視覺、機(jī)器人導(dǎo)航和定位、模式識(shí)別、多傳感器融合及人機(jī)接口等，涉及機(jī)械、控制、傳感器、人工智能等技術(shù)。智能輪椅主要由導(dǎo)航系統(tǒng),運(yùn)動(dòng)控制系統(tǒng)和人機(jī)接口三部分組成。事實(shí)上，智能輪椅的智能更多地體現(xiàn)在輪椅與使用者之間和諧的交互上。友好的人機(jī)接口已成為實(shí)現(xiàn)智能輪椅主要功能的關(guān)鍵。

鑒于人機(jī)接口的重要性，世界各地的研究者對(duì)智能輪椅的人機(jī)接口進(jìn)行了深入的研究?；谏眢w運(yùn)動(dòng)[1],頭部運(yùn)動(dòng)[2]，觸摸屏[3]，語音[4]，呼吸運(yùn)動(dòng)[5]，計(jì)算機(jī)視覺和生物電信號(hào)等的人機(jī)接口已經(jīng)進(jìn)入人們的視野。其中基于計(jì)算機(jī)視覺的人機(jī)接口又可以細(xì)分為基于手部姿態(tài)[6]，頭部姿態(tài)[7]，眼睛凝視[8]，口形[9]，臉部方向[10]等的人機(jī)接口。由于基于單獨(dú)特征的人機(jī)接口往往具有局限性，所以在實(shí)際應(yīng)用中往往會(huì)采用基于多種特征組合的人機(jī)接口[8-9][11]。

肌電信號(hào)在實(shí)際生活中已經(jīng)得到了廣泛應(yīng)用。肌電信號(hào)已經(jīng)被研究用作殘障病人義肢的控制信號(hào)；另外肌電信號(hào)中包含了豐富的人體信息，比如人體動(dòng)作的力量信息和位置信息等。相對(duì)于傳統(tǒng)的人機(jī)接口技術(shù)，肌電信號(hào)能直接反映人的動(dòng)作意圖，其受到的外界干擾更少，輪椅操作的可靠性、安全性和人體舒適性更高。目前針對(duì)智能輪椅中基于肌電信號(hào)的人機(jī)接口的研究主要集中在如何從肌電信號(hào)中提取相關(guān)信息并將之轉(zhuǎn)換為相應(yīng)的輪椅操控命令。

1基于肌電信號(hào)的人機(jī)接口系統(tǒng)的主要工作

人在做肢體運(yùn)動(dòng)時(shí)，肢體的特定關(guān)節(jié)運(yùn)動(dòng)由其對(duì)應(yīng)的肌肉群控制，由相應(yīng)肌群運(yùn)動(dòng)檢測(cè)到的肌電信號(hào)不但能反映關(guān)節(jié)的伸屈狀態(tài)和伸屈強(qiáng)度，還能反映動(dòng)作完成過程中肢體的形狀、位置和運(yùn)動(dòng)等信息，這就是采用肌電信號(hào)識(shí)別人體動(dòng)作的生理學(xué)依據(jù)。基于肌電信號(hào)的人機(jī)接口系統(tǒng)的主要工作依次包括信號(hào)采集，信號(hào)調(diào)理，A/D轉(zhuǎn)換，活動(dòng)段檢測(cè)，特征提取，分類識(shí)別，控制指令轉(zhuǎn)換，D/A轉(zhuǎn)換，輪椅驅(qū)動(dòng)。其工作重點(diǎn)在于信號(hào)調(diào)理，活動(dòng)段檢測(cè)，特征提取和分類識(shí)別。

1.1肌電信號(hào)的采集

根據(jù)測(cè)量中所用的引導(dǎo)電極和安放位置的不同，肌電信號(hào)可分為針電極肌電信號(hào)NEMG (Needle EMG)和表面電極肌電信號(hào)SEMG (Surface EMG)兩種，前者是以針形電極（如圖1所示）為引導(dǎo)電極，后者則是以表面電極（如圖2所示）為引導(dǎo)電極，將其安置在皮膚表面時(shí)測(cè)量到的肌肉活動(dòng)在檢測(cè)表面處的電位綜合。針電極插入人體時(shí)會(huì)對(duì)肌肉和脂肪組織造成損傷，不宜反復(fù)多次或過長(zhǎng)時(shí)間測(cè)量，也不宜同時(shí)測(cè)量多路信號(hào)。所以一般采用表面電極來測(cè)量，其最大優(yōu)點(diǎn)是測(cè)量的無損傷性。表面電極具有較大的檢測(cè)表面和較低的空間分辨率，其所記錄的信號(hào)為一定范圍內(nèi)肌纖維電活動(dòng)的總和。一般SEMG的幅度范圍為0-1.5mV，帶寬為0.5-2kHz，主要能量集中在50-150Hz范圍內(nèi)。

2結(jié)束語

肌電信號(hào)在智能輪椅中的研究已經(jīng)取得了一定的成績(jī)，但很多方面仍然需要改進(jìn)和創(chuàng)新以達(dá)到更好的效果，我們可以從以下方面考慮：

1)改進(jìn)信號(hào)的提取方法，最大限度提取人機(jī)交互所需的有用信息，提高源信號(hào)的提取精度和可靠性。

2)由于肌電信號(hào)本身極其微弱，非常容易受到各種干擾與噪聲的影響，因此必須解決肌電信號(hào)的預(yù)處理問題。

3)改進(jìn)特征提取與模式識(shí)別的算法，在確保安全性和高識(shí)別準(zhǔn)確率的情況下，盡量提高信息傳輸率和系統(tǒng)實(shí)時(shí)性。

4)提高人機(jī)接口系統(tǒng)的適應(yīng)能力：由于人的個(gè)體差異較大，因此需要研究適應(yīng)對(duì)象特點(diǎn)的特征選擇算法。

參考文獻(xiàn)：

[1] Yokota S.Hashimoto H,Ohyama Y,et al.The electric wheelchair controlled by human body motion[C].Human System Interactions,HSI 09. 2nd Conference on, 2009,23( 21):247-250.

[2] Kupetz D.J,Wentzell S A,Bu Sha B F. Head motion controlled power wheelchair[C].Bioengineering Conference, Proceedings of the 2010 IEEE 36th Annual Northeast ,2010,2(1):26-28.

[3] Montesano L,Diaz M,Bhaskar S,et al. Towards an Intelligent Wheelchair System for Users With Cerebral Palsy[J].Neural Systems and Rehabilitation Engineering. IEEE Transactions on, 2010,2(18):193-202.

[4] Nishimori M,Saitoh T,Konishi R.Voice controlled intelligent wheelchair[C].SICE,Annual Conference, 2007,340(336):17-20.

[5] Mazo M.An integral system for assisted mobility [automated wheelchair[J].Robotics & Automation Magazine, IEEE,2001,1(8):46-56.

[6] Posada-Gomez R,Sanchez-Medel L H,Hernandez G A,et al. A Hands Gesture System Of Control For An Intelligent Wheelchair[C].Electrical and Electronics Engineering, ICEEE 2007. 4th International Conference on,2007,71(68): 5-7.

[7] Zhang-fang Hu,Lin Li,Yuan Luo,et al. A novel intelligent wheelchair control approach based on head gesture recognition[C].Computer Application and System Modeling,ICCASM,2010 International Conference on,2010, V6-159-V6-163 (6) :22-24 .

[8] Matsumoto Y,Zelinsky A.An algorithm for real-time stereo vision implementation of head pose and gaze direction measurement[C].Automatic Face and Gesture Recognition,2000. Proceedings . Fourth IEEE International Conference on,2000:499-504.

[9] Jin Sun Ju,Yun Hee Shin, Eun Yi Kim,et al. Intelligent wheelchair using face and mouth shape recognition[C].Consumer Electronics,2008. ICCE 2008.Digest of Technical Papers.International Conference on, 2008,2(1):9-13.

[10] Kuno Y,Murashina T,Shimada N,et al. Intelligent wheelchair remotely controlled by interactive gestures[C].Pattern Recognition,2000. Proceedings.15th International Conference on, 2000(4):672-675.

[11] Feng Wang, Xiangshi Ren, A Survey of Human Computer Interaction Models for the Disabled[J]. Information-An International Interdisciplinary Journal,2010,3(10):585-591.

[12] Kyuwan Choi,Sato M,Koike Y.A new, human-centered wheelchair system controlled by the EMG signal[C].Neural Networks,IJCNN 06. International Joint Conference on, 2010:4664-4671.

[13] Inhyuk Moon,Myungjoon Lee,Junuk Chu,et al.Wearable EMG-based HCI for Electric-Powered Wheelchair Users with Motor Disabilities[C].Robotics and Automation, ICRA 2005.Proceedings of the 2005 IEEE International Conference on,2005,2654(2649): 18-22.

[14] Jeong-Su Han, et al. Development of an EMG-based Powered Wheelchair Controller for Users with High-level Spinal Cord Injury[C]. Proc. of Int. Conf. on Control,Automation and Systems, 2001:6-503.

[15] Lai Wei,Huosheng Hu,Kui Yuan.Use of forehead bio-signals for controlling an Intelligent Wheelchair[C]. Robotics and Biomimetics,ROBIO 2008.IEEE International Conference on,2009,113(108): 22-25.

[16] Hashimoto M,Takahashi K,Shimada M.Wheelchair control using an EOG- and EMG-based gesture interface[C].Advanced Intelligent Mechatronics,AIM 2009.IEEE/ASME International Conference on, 2009 ,1217(1212):14-17.