CAJ | 학술논문

文中提出一种基于独立成分分析的非监督气体模式识别方法,用多元统计理论中的独立成分分析(independent component analysis-ICA)来分析金属氧化物半导体气体传感器阵列响应数据,进而实现对不同种类气体的分类.对所设计的电子鼻实验系统测量得到的气体传感器阵列稳态响应数据进行了白化和快速定点独立成分分析(FastICA)处理.实验结果表明该方法对区分一氧化碳(CO)、甲烷(CH_4)和氢气(H_2)3种气体有很高的识别率.
문중제출일충기우독립성분분석적비감독기체모식식별방법,용다원통계이론중적독립성분분석(independent component analysis-ICA)래분석금속양화물반도체기체전감기진렬향응수거,진이실현대불동충류기체적분류.대소설계적전자비실험계통측량득도적기체전감기진렬은태향응수거진행료백화화쾌속정점독립성분분석(FastICA)처리.실험결과표명해방법대구분일양화탄(CO)、갑완(CH_4)화경기(H_2)3충기체유흔고적식별솔.
An ICA-based unsupervised pattern recognition method was proposed. Independent component analysis (ICA) as a multivariate statistical tool was used to analyze the steady-state responses generated by an array of commercial metal-oxide gas sensors in an electronic nose system and to classify the different gases. Whitening and FastICA iteration are used to process the gas sensor array measurement data obtained by an experimental electronic nose system. The technique is demonstrated by discriminating three gases carbon monoxide (CO), methane (CH_4) and hydrogen (H_2) with a high success rate.