本文選題：太赫茲 + 太赫茲時(shí)域光譜　； 參考：《國(guó)防科學(xué)技術(shù)大學(xué)》2015年碩士論文

【摘要】：太赫茲波是一種頻率范圍一般在0.1THz~10THz之間的電磁波,介于微波和紅外之間。由于20世紀(jì)80年代之前由于缺乏大功率的太赫茲源和高靈敏度的探測(cè)器,這一波段一直沒(méi)有得到有效的開(kāi)發(fā)和利用,因此被稱(chēng)為太赫茲間隙(THz gap)。由于太赫茲波具有很多優(yōu)良的性質(zhì),并且隨著太赫茲源和探測(cè)器技術(shù)的提高,以太赫茲時(shí)域光譜技術(shù)為代表的太赫茲?rùn)z測(cè)技術(shù)得到了廣泛的應(yīng)用。本文利用太赫茲時(shí)域光譜技術(shù)開(kāi)展材料辨識(shí)技術(shù)研究,主要研究?jī)?nèi)容和成果如下:(1)針對(duì)傳統(tǒng)的光譜平滑方法,參數(shù)無(wú)法自動(dòng)設(shè)置,容易出現(xiàn)過(guò)平滑,使得光譜中的吸收峰被減弱甚至消除,同時(shí)具有無(wú)法去除脈沖噪聲的缺點(diǎn),提出了一種改進(jìn)的分?jǐn)?shù)階積分光譜平滑方法,首先使用中值濾波器去除脈沖噪聲,然后對(duì)分?jǐn)?shù)階積分平滑后的光譜進(jìn)行加權(quán)平均,提高了算法對(duì)含多種噪聲類(lèi)型光譜的適應(yīng)性,能夠取得更好的平滑效果和更高的信噪比。(2)針對(duì)太赫茲光譜中存在的基線(xiàn)漂移現(xiàn)象,提出了一種非對(duì)稱(chēng)ε-不敏感支持向量機(jī),通過(guò)對(duì)正的擬合誤差和負(fù)的擬合誤差使用不同程度的損失函數(shù),使得擬合的曲線(xiàn)不再是針對(duì)光譜本身而是光譜基線(xiàn),用原始光譜減去擬合的基線(xiàn)從而對(duì)光譜實(shí)現(xiàn)了基線(xiàn)校正,相對(duì)于傳統(tǒng)方法,校正的結(jié)果可以更好的保持吸收峰的原始形狀。(3)針對(duì)傳統(tǒng)的太赫茲譜分類(lèi)識(shí)別方法將太赫茲光譜單純的作為一個(gè)普通一維向量,沒(méi)有依據(jù)太赫茲光譜本身的特點(diǎn)來(lái)處理,導(dǎo)致識(shí)別率不高的現(xiàn)象,提出了一種基于多尺度分析的特征選擇方法,能夠有效的提取光譜特征吸收峰,去除噪聲帶來(lái)的干擾。同時(shí)針對(duì)支持向量機(jī)訓(xùn)練多類(lèi)別、大規(guī)模數(shù)據(jù)時(shí),內(nèi)存消耗和時(shí)間消耗過(guò)大的缺點(diǎn),提出了一種改進(jìn)的DAG-SVM快速算法,該算法首先使用基于均值漂移的支持向量預(yù)選取方法,避免了支持向量漏選的現(xiàn)象,然后根據(jù)類(lèi)別之間的鄰接圖對(duì)原始DAG-SVM算法的決策樹(shù)進(jìn)行修剪,能夠減少訓(xùn)練器的個(gè)數(shù),加快訓(xùn)練的速度。
[Abstract]:Terahertz wave is a kind of electromagnetic wave with frequency range between 0.1THz~10THz and infrared. Due to the lack of high power terahertz source and high sensitivity detector before 1980s, this band has not been developed and utilized effectively, so it is called THz gap. With the improvement of terahertz source and detector technology, terahertz detection technology, represented by terahertz time-domain spectroscopy, has been widely used. In this paper, terahertz time-domain spectroscopy is used to develop material identification technology. The main research contents and results are as follows: (1) aiming at the traditional spectral smoothing method, the parameters can not be set automatically, and it is easy to occur smoothing. The absorption peak in the spectrum is weakened or eliminated, and the pulse noise can not be removed. An improved fractional-order integral spectral smoothing method is proposed. First, the median filter is used to remove the pulse noise. Then, the weighted average of the spectrum after fractional integral smoothing is carried out, which improves the adaptability of the algorithm to the spectrum with multiple noise types. A novel asymmetric 蔚 -insensitive support vector machine is proposed to solve the baseline drift phenomenon in terahertz spectrum, which can achieve better smoothing effect and higher signal-to-noise ratio (SNR). By using different loss functions to the positive and negative fitting errors, the fitted curve is no longer aimed at the spectrum itself but the spectral baseline. The baseline correction is realized by subtracting the fitted baseline from the original spectrum. Compared with the traditional method, the corrected results can better maintain the original shape of the absorption peak.) the terahertz spectrum is simply regarded as a general one-dimensional vector for the traditional THz spectrum classification and recognition method. A new feature selection method based on multi-scale analysis is proposed, which can extract spectral characteristic absorption peak effectively and remove the interference caused by noise because it is not dealt with according to the characteristics of terahertz spectrum itself, which leads to low recognition rate. At the same time, an improved DAG-SVM fast algorithm is proposed to solve the problem of excessive memory and time consumption when SVM is training multi-class and large-scale data. Firstly, support vector pre-selection method based on mean shift is used in this algorithm. The missing support vector is avoided and the decision tree of the original DAG-SVM algorithm is pruned according to the adjacent graph between classes, which can reduce the number of trainers and speed up the training.
【學(xué)位授予單位】：國(guó)防科學(xué)技術(shù)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類(lèi)號(hào)】：TB30;O441

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 侯培國(guó);李寧;常江;王書(shū)濤;宋濤;;SG平滑和IBPLS聯(lián)合優(yōu)化水中油分析方法的研究[J];光譜學(xué)與光譜分析;2015年06期

2 盧承振;劉維;孫萍;王奕新;江兆鳳;;不同水的太赫茲時(shí)域光譜[J];激光與光電子學(xué)進(jìn)展;2015年04期

3 李向軍;楊曉杰;劉建軍;;基于反射式太赫茲時(shí)域譜的水太赫茲光學(xué)參數(shù)測(cè)量與誤差分析[J];光電子·激光;2015年01期

4 于東鈺;楊鴻儒;李宏光;;太赫茲微弱信號(hào)檢測(cè)系統(tǒng)噪聲處理方法研究[J];應(yīng)用光學(xué);2012年06期

5 蔣方純;;機(jī)器學(xué)習(xí)應(yīng)用中特征缺失研究[J];深圳信息職業(yè)技術(shù)學(xué)院學(xué)報(bào);2012年03期

6 梅玲;張亞麗;;結(jié)構(gòu)化數(shù)據(jù)核函數(shù)綜述[J];計(jì)算機(jī)工程;2010年23期

7 黃為軍;柴玉澤;;探索太赫茲——專(zhuān)訪(fǎng)首都師范大學(xué)太赫茲實(shí)驗(yàn)室主任張存林[J];科學(xué)中國(guó)人;2010年11期

8 肖健;高愛(ài)華;;光電導(dǎo)天線(xiàn)產(chǎn)生太赫茲波的研究[J];應(yīng)用光學(xué);2010年03期

9 張存林;牧凱軍;;太赫茲波譜與成像[J];激光與光電子學(xué)進(jìn)展;2010年02期

10 李鐵元;婁采云;王黎;黃縉;趙國(guó)忠;石小溪;;低溫生長(zhǎng)砷化鎵光電導(dǎo)天線(xiàn)產(chǎn)生太赫茲波[J];中國(guó)激光;2009年04期

相關(guān)博士學(xué)位論文 前2條

1 李萌;太赫茲時(shí)域光譜品質(zhì)影響因素及改進(jìn)研究[D];天津大學(xué);2011年

2 蒲亦非;分?jǐn)?shù)階微積分在現(xiàn)代信號(hào)分析與處理中應(yīng)用的研究[D];四川大學(xué);2006年

相關(guān)碩士學(xué)位論文 前3條

1 黃時(shí)光;太赫茲波傳輸特性研究[D];西安電子科技大學(xué);2010年

2 王卓;光學(xué)方法產(chǎn)生連續(xù)、準(zhǔn)連續(xù)THz波的理論研究[D];天津大學(xué);2007年

3 徐新龍;從THz時(shí)間波形中提取材料參數(shù)的方法和分析[D];首都師范大學(xué);2003年

，

本文編號(hào)：1978095