本文關(guān)鍵詞：機(jī)械狀態(tài)流形特征增強(qiáng)理論及監(jiān)測(cè)診斷方法研究　出處：《中國(guó)科學(xué)技術(shù)大學(xué)》2017年博士論文　論文類(lèi)型：學(xué)位論文

  更多相關(guān)文章： 流形特征增強(qiáng) 特征信息再學(xué)習(xí) 參照化流形 深度化流形 稀疏化流形 機(jī)械設(shè)備 狀態(tài)監(jiān)測(cè)與診斷

【摘要】：本論文以機(jī)械設(shè)備狀態(tài)的精確辨識(shí)與故障的有效診斷為研究目標(biāo),著眼于流形學(xué)習(xí)對(duì)本質(zhì)特征信息上的挖掘能力,通過(guò)分析基于狀態(tài)維護(hù)的狀態(tài)特征、模式特征、信號(hào)特征三種特征信息的特點(diǎn),在理論上提出了參照化流形、深度化流形以及稀疏化流形三種流形增強(qiáng)學(xué)習(xí)新模式,在內(nèi)容實(shí)現(xiàn)了對(duì)狀態(tài)差異特征、模式敏感特征、信號(hào)模態(tài)特征的流形增強(qiáng)學(xué)習(xí),建立了一種流形特征增強(qiáng)學(xué)習(xí)的設(shè)備狀態(tài)監(jiān)測(cè)與故障診斷的研究體系,并分別就上述三種流形增強(qiáng)學(xué)習(xí)的若干理論和問(wèn)題進(jìn)行了深入研究。在設(shè)備狀態(tài)特征增強(qiáng)學(xué)習(xí)中,基于機(jī)械設(shè)備具有長(zhǎng)期健康運(yùn)行、非健康狀態(tài)同健康狀態(tài)存在著相對(duì)較大差異的物理特點(diǎn),對(duì)監(jiān)測(cè)樣本建立以相同數(shù)據(jù)為參照樣本的實(shí)時(shí)比對(duì)模型,利用流形學(xué)習(xí)獲取該模型參照化流形特征空間,通過(guò)分析空間聚類(lèi)遷移變化來(lái)刻畫(huà)設(shè)備狀態(tài)的退化情況。該參照化流形空間聚類(lèi)分布有效地突出了監(jiān)測(cè)狀態(tài)相比于健康狀態(tài)的變化信息,揭露了設(shè)備狀態(tài)的差異特征。基于狀態(tài)特征構(gòu)造方式和流形空間的不同,分別發(fā)展了多元統(tǒng)計(jì)特征的流形空間聚類(lèi)和無(wú)特征的拉斯曼流形流形基空間聚類(lèi)兩種狀態(tài)監(jiān)測(cè)方案。并進(jìn)一步提出了參照化流形空間聚類(lèi)控制圖,實(shí)現(xiàn)對(duì)設(shè)備狀態(tài)退化的多階段量化分析以及早期退化時(shí)刻的放大和報(bào)警。在故障模式特征增強(qiáng)學(xué)習(xí)中,基于參照化流形對(duì)于差異性特征增強(qiáng)的作用,建立多單元比對(duì)模型對(duì)原始特征形成擴(kuò)展學(xué)習(xí)、進(jìn)一步利用多層流形學(xué)習(xí)實(shí)現(xiàn)對(duì)特征的級(jí)聯(lián)式多層學(xué)習(xí)。通過(guò)這種多層化學(xué)習(xí)方式,建立了如同多層神經(jīng)網(wǎng)絡(luò)特征學(xué)習(xí)模式,在這種特征擴(kuò)展再學(xué)習(xí)過(guò)程中完成對(duì)模式敏感特征的增強(qiáng)學(xué)習(xí)。本文利用這種級(jí)聯(lián)式深度化流形學(xué)習(xí)方式,提出了多比對(duì)模型的兩層流形特征增強(qiáng)算法,有效的增強(qiáng)了不同故障類(lèi)型之間的差異性、增強(qiáng)了模式特征敏感性,對(duì)于實(shí)際故障識(shí)別顯示了與傳統(tǒng)的流形特征識(shí)別方法的優(yōu)越性。在信號(hào)模態(tài)特征增強(qiáng)學(xué)習(xí)中,利用時(shí)頻流形能夠提取瞬態(tài)特征本質(zhì)信息的優(yōu)勢(shì),引入稀疏分析原理,完成兩者優(yōu)勢(shì)互補(bǔ),重新建立一種稀疏化流形特征學(xué)習(xí)新模式,實(shí)現(xiàn)對(duì)信號(hào)稀疏模態(tài)特征的深度挖掘與再學(xué)習(xí),克服傳統(tǒng)時(shí)頻流形學(xué)習(xí)在原始瞬態(tài)特征失真以及無(wú)法去除強(qiáng)噪聲的缺陷。基于這種稀疏化流形分析的思想,分別提出了時(shí)頻匹配稀疏和包絡(luò)移不變稀疏的流形模態(tài)特征增強(qiáng)方法,實(shí)現(xiàn)了對(duì)原始信號(hào)波形特征的保持與恢復(fù),有利于精確的故障模態(tài)診斷研究。綜上所述,本論文研究了流形特征增強(qiáng)學(xué)習(xí)在CBM中的理論方法,包括狀態(tài)監(jiān)測(cè)預(yù)警、故障模式識(shí)別、故障信號(hào)診斷三個(gè)方面,提出了參照化流形、深度化流形以及稀疏化流形特征增強(qiáng)學(xué)習(xí)方法,進(jìn)行了一次較為完整的特征增強(qiáng)學(xué)習(xí)的研究。和傳統(tǒng)方法對(duì)比,流形特征增強(qiáng)可以實(shí)現(xiàn)更有效、更精準(zhǔn)、更敏感的狀態(tài)監(jiān)測(cè)與故障診斷分析,對(duì)系統(tǒng)化的機(jī)械設(shè)備CBM研究具有重要意義。
[Abstract]:In this paper, the effective diagnosis accurate identification and fault state of the machinery and equipment as the research object, focusing on the manifold learning the essential characteristics of information mining capacity, through the analysis of characteristics, state maintenance based on mode characteristics, characteristics of signal characteristics of three kinds of feature information, in theory, puts forward a reference of the depth of the manifold manifold. And the sparse manifold three manifold reinforcement learning mode, characteristics of the state is realized by a difference in content, pattern sensitive features, manifold signal modal characteristics of reinforcement learning, a manifold learning feature enhancement system of state monitoring and fault diagnosis of equipment, and these three manifolds enhanced several theories and problems learning is studied. In the equipment characteristics of reinforcement learning, mechanical equipment has a long-term healthy operation based on non health status with the healthy state of existence Relatively large differences in physical characteristics, to establish monitoring samples with the same data for real-time alignment model reference samples, learning to obtain the reference models of manifold feature space using manifold, through the degradation of spatial clustering analysis to characterize the change of equipment condition. According to the distribution of cluster manifold space effectively highlights the change of state monitoring information compared to the healthy state, reveals the differences of equipment condition features. Feature structure and manifold space based on the different were the development of a manifold space clustering multivariate statistical feature and non feature based spatial clustering Larsemann manifold two state monitoring scheme. And put forward the reference of manifold space clustering control chart, realization the degradation of device status and quantitative analysis of multi stage amplification and early degradation time alarm. In fault pattern feature enhancement study According to the study, to enhance the diversity of manifold features based on the original features of the formation of an extended learning based multi unit alignment model, further use of manifold learning to achieve multilayer cascade multilayer characteristics of learning. Through this kind of multilayer learning methods, such as the establishment of a multilayer neural network with the characteristics of learning mode, in this feature expansion again learn to complete the reinforcement learning pattern of sensitive features in the process. Using this kind of cascade depth of manifold learning method, and puts forward two layers of the model than the manifold feature enhancement algorithm, the difference between the effective enhancement of different fault types, enhanced pattern features of sensitivity for the actual fault identification shows the superiority and manifold features the traditional recognition methods. In signal modal characteristics of reinforcement learning, using time-frequency manifold can extract the transient characteristics of essential information advantage, Introduce the principle of sparse analysis, complete complementary advantage, the re establishment of a new model of sparse manifold learning characteristics, implementation and study depth of excavation of sparse modal characteristics, to overcome the traditional time-frequency manifold learning defects in the original distortion transient characteristics and unable to remove the strong noise. The analysis of this sparse manifold based on the idea. When the frequency matching are proposed and sparse envelope method to enhance shift invariant sparse modal manifold, implement maintain and restore the original signal waveform characteristics, which is helpful to the study of modal fault diagnosis accurately. To sum up, this paper studies the characteristics of reinforcement learning theory of manifold method in CBM, including the state monitoring and early warning fault pattern recognition, fault diagnosis, signal three aspects, put forward the reference of the depth of the manifold, manifold and manifold feature sparse reinforcement learning method, in a A relatively complete research on feature enhanced learning. Compared with traditional methods, manifold feature enhancement can achieve more effective, more precise and sensitive state monitoring and fault diagnosis analysis, which is of great significance for the research of systematic mechanical equipment CBM.

【學(xué)位授予單位】：中國(guó)科學(xué)技術(shù)大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：TH17

【相似文獻(xiàn)】

相關(guān)期刊論文 前8條

1 焦巍;劉光斌;張金生;張博;喬玉坤;;基于免疫粒子群算法的地磁特征區(qū)域選擇[J];宇航學(xué)報(bào);2010年06期

2 劉波,方舟,張君安;圖像特征區(qū)域質(zhì)心位移的檢測(cè)[J];西安工業(yè)學(xué)院學(xué)報(bào);2004年02期

3 劉復(fù)友;杜嶺;;安徽省村莊特征區(qū)域劃分與優(yōu)化策略研究[J];池州學(xué)院學(xué)報(bào);2011年06期

4 孫劉杰;李晨璐;;基于圖像特征區(qū)域的加密全息水印算法[J];包裝工程;2013年15期

5 張薇;;一種新的基于特征區(qū)域的圖像配準(zhǔn)算法[J];電腦知識(shí)與技術(shù);2006年17期

6 郝國(guó)棟;程筱勝;戴寧;俞青;;基于形態(tài)學(xué)的牙齒模型交互分割[J];中國(guó)制造業(yè)信息化;2008年01期

7 李麗麗;孫勁光;;SIFT特征區(qū)域Contourlet域遙感圖像水印算法[J];信號(hào)處理;2012年09期

8 王聃;李浩茹;鄧春偉;;基于局部區(qū)域LBP特征的人臉識(shí)別研究[J];黑龍江工程學(xué)院學(xué)報(bào)(自然科學(xué)版);2012年03期

相關(guān)會(huì)議論文 前4條

1 蒙山;張有為;;人機(jī)自然交互過(guò)程中的人臉檢測(cè)及其特征區(qū)域的定位[A];圖像 仿真 信息技術(shù)——第二屆聯(lián)合學(xué)術(shù)會(huì)議論文集[C];2002年

2 王清;姜志國(guó);趙丹培;安萌;;一種基于特征區(qū)域的空間有形目標(biāo)識(shí)別方法[A];第八屆全國(guó)信號(hào)與信息處理聯(lián)合學(xué)術(shù)會(huì)議論文集[C];2009年

3 孔維武;楊立瑞;;雙能X射線包裹圖像特征區(qū)域剔除方法[A];第十五屆全國(guó)圖象圖形學(xué)學(xué)術(shù)會(huì)議論文集[C];2010年

4 徐昊;黃峻;江文文;周小虎;;武鄂黃城鎮(zhèn)連綿帶規(guī)劃建設(shè)思路研究[A];轉(zhuǎn)型與重構(gòu)——2011中國(guó)城市規(guī)劃年會(huì)論文集[C];2011年

相關(guān)博士學(xué)位論文 前3條

1 丁曉喜;機(jī)械狀態(tài)流形特征增強(qiáng)理論及監(jiān)測(cè)診斷方法研究[D];中國(guó)科學(xué)技術(shù)大學(xué);2017年

2 樂(lè)靜;基于坐標(biāo)和圖像技術(shù)的三維表面特征區(qū)域的檢測(cè)[D];西安理工大學(xué);2007年

3 杜耀華;基因組轉(zhuǎn)錄調(diào)控元件的特征分析與識(shí)別算法研究[D];國(guó)防科學(xué)技術(shù)大學(xué);2006年

相關(guān)碩士學(xué)位論文 前5條

1 梅晶;利用密集特征進(jìn)行目標(biāo)自適應(yīng)跟蹤算法研究[D];江西理工大學(xué);2015年

2 尹廣學(xué);面向生物敏感性的魯棒可逆水印技術(shù)研究[D];西安電子科技大學(xué);2015年

3 萬(wàn)麗;球形立體視覺(jué)特征區(qū)域提取及匹配方法研究[D];天津理工大學(xué);2016年

4 李磊;一種基于特征提取的人臉圖片加密算法的設(shè)計(jì)與實(shí)現(xiàn)[D];華中科技大學(xué);2015年

5 陳亮;一種基于特征與極性映射方法的觀點(diǎn)分析系統(tǒng)的研究與設(shè)計(jì)[D];東北大學(xué);2010年

，

本文編號(hào)：1390921