結合雲端穿戴訓練裝置之腦波人機介面-子計畫一：結合穿戴訓練裝置之人機介面腦波辨識技術開發;Development of Brain-Wave Recognition Technique Using Wearable Training Device

NCUIR > College of Electrical Engineering & Computer Science > Department of Electrical Engineering > Research Project > Item 987654321/78766

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/78766

Title:	結合雲端穿戴訓練裝置之腦波人機介面-子計畫一：結合穿戴訓練裝置之人機介面腦波辨識技術開發;Development of Brain-Wave Recognition Technique Using Wearable Training Device
Authors:	李柏磊;李偉強
Contributors:	國立中央大學電機工程學系
Keywords:	腦波人機介面;機械外骨骼系統;全息希爾伯特譜分析;事件相關非同步/同步分析;Brain computer interface (BCI);Robotic exoskeleton system;Holo-Hilbert spectral analysis (HHSA);Event-related desynchronization / synchronization (ERD/ERS)
Date:	2018-12-19
Issue Date:	2018-12-20 13:47:26 (UTC+8)
Publisher:	科技部
Abstract:	隨著老齡化社會的來臨，老年與癱瘓人口數目逐年攀升，也讓行動輔具的需求與日遽增。本計畫的目的在於開發一種新型的腦波控制機械骨骼系統，此機械骨骼系統將直接由腦波控制，不需透過手動操作或是肌電波，可以適用於重度癱瘓或是全癱瘓病人使用。然而，腦波人機介面最難的部分在於辨識各種不同運動引起的腦波型態，因此本子計畫結合穿戴式動作感測器與腦電波，開發穿戴式感測器時間標記腦波分析技術，讓使用者在生活中標記不同運動的腦波，然後由全息希爾伯特譜分析與事件相關非同步/同步訊號獲得信號特徵，最後由可適性模糊神經網路與隱藏式馬可夫鍊進行訊號即時辨識。本系統的開發將有助於減輕社會照護成本，提供癱瘓病人行動與溝通能力。 ;With the coming of aging society, the human population of aged and paralyzed people keeps rising which results in the increase of the need for mobility assistive devices. The aim of this project intends to develop a novel brain-controlled robtic exoskeleton system. The system is controlled directly through brain waves, independent of hand manipulation and muscle activities, which is suitable to be used by heavily paralyzed or totally paralyzed patients. However, the most difficult part in designing a brain computer interface is the recognition of brain-wave patterns induced by different kinds of movements. To solve this difficulty, we intend to combine wearable motion detector and electroencephalography (EEG) to develop motion-marked EEG analysis technique. Users’ daily-life movement events in daily life will be marked and used as trigger events for EEG analysis. Then, event-marked EEG signals will be analyzed by Holo-Hilbert spectral analysis (HHSA) and evenet-related desynchronizaion / synchronization (ERD/ERS) to detect user’s motion intention, and the motion intention of different limbs will be recognized by adaptive neuron-fuzzy classifier (ANFC) and hidden markov model (HMM) to achieve real-time motion recognition. The development of this system can help the alleviation of social health-care cost and provide a channel for paralyzed patient to move and communicate with external environments.
Relation:	財團法人國家實驗研究院科技政策研究與資訊中心
Appears in Collections:	[Department of Electrical Engineering] Research Project

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	228	View/Open

社群 sharing

Loading...