本文關(guān)鍵詞： 過程監(jiān)測 非線性故障檢測 核方法 流形學(xué)習(xí)　出處：《浙江大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：過程監(jiān)測系統(tǒng)能夠?qū)崟r地監(jiān)測生產(chǎn)過程,在保障工況平穩(wěn)運行、改善產(chǎn)品質(zhì)量及降低能耗等方面越來越發(fā)揮著不可替代的作用。大數(shù)據(jù)時代正在隨著信息化程度不斷發(fā)展以及硬件存儲和計算水平的不斷提升而到來,數(shù)據(jù)的極大豐富使得基于數(shù)據(jù)驅(qū)動的過程監(jiān)測方法成為了近年來研究的熱點。相應(yīng)地,這些豐富的數(shù)據(jù)也使得基于數(shù)據(jù)驅(qū)動的過程監(jiān)測方法面臨著更多的挑戰(zhàn)。本文針對工業(yè)生產(chǎn)過程當(dāng)中的非線性特性,采用核方法和流形學(xué)習(xí)(Manifold Learning)提出了兩種基于核熵成分分析(Kernel Entropy Component Analysis,KECA)的非線性故障檢測方法。具體研究內(nèi)容包括:(1)針對單一模型的KECA方法并不能夠有效地檢測工業(yè)過程當(dāng)中存在的不同類型的故障的問題,提出基于集成學(xué)習(xí)和貝葉斯推論的改進KECA故障檢測方法。由于不同類型的故障往往需要不同大小的核參數(shù)使得其具有較好的檢測效果,本文采用相同的訓(xùn)練數(shù)據(jù)對不同核參數(shù)構(gòu)造的KECA模型進行訓(xùn)練實現(xiàn)離線建模。在建立模型之后通過貝葉斯推論將這些模型的在線檢測效果轉(zhuǎn)化為概率的形式,最后將這些概率形式的檢測結(jié)果根據(jù)加權(quán)方式組合形成一個最終的檢測結(jié)果,給予對特定故障有較好檢測效果的模型較大的權(quán)重,從而實現(xiàn)了對不同故障類型均具有較好的檢測效果的目的。(2)考慮到KECA能夠更全面地選擇降維過程中數(shù)據(jù)的投影方向的優(yōu)點,本文將信息熵的思想引入到最大方差展開(Maximum Variance Unfolding,MVU)當(dāng)中,提出了一種基于KECA-MVU的非線性故障檢測方法。利用瑞利熵來衡量由MVU學(xué)習(xí)得到的核矩陣經(jīng)過數(shù)據(jù)變換之后的信息保留有效程度,根據(jù)瑞利熵最大的前幾項所對應(yīng)的特征向量作為數(shù)據(jù)的投影方向,實現(xiàn)了數(shù)據(jù)的有效壓縮。最后采用線性回歸的方法估計了輸入數(shù)據(jù)到低維數(shù)結(jié)構(gòu)的最優(yōu)投影矩陣,由該投影矩陣實現(xiàn)過程中故障的在線檢測。最后,總結(jié)了本文的主要研究成果,并闡述了未來研究工作的難點及展望。
[Abstract]:The process monitoring system can monitor the production process in real time and run smoothly in the guaranteed working conditions. Improving the quality of products and reducing energy consumption are playing an irreplaceable role. Big data era is coming along with the development of information technology and the continuous improvement of hardware storage and computing. With the abundance of data, the data-driven process monitoring method has become a hot topic in recent years. These abundant data also make the data-driven process monitoring method face more challenges. Using kernel method and Manifold learning (Manifold learning), two methods based on kernel entropy component analysis are proposed. Kernel Entropy Component Analysis. The specific research contents include: 1) the KECA method for a single model can not effectively detect the problems of different types of faults in the industrial process. An improved KECA fault detection method based on ensemble learning and Bayesian inference is proposed. Because different types of faults often require different kernel parameters, it has better detection effect. In this paper, we use the same training data to train KECA models with different kernel parameters to realize off-line modeling. After establishing the model, the on-line detection effect of these models is transformed into probabilistic form by Bayesian inference. Style. Finally, these probabilistic detection results are combined according to the weighted method to form a final detection result, which gives a larger weight to the model with better detection effect for a particular fault. Therefore, the purpose of better detection effect for different fault types is realized. (2) considering the advantage that KECA can more comprehensively select the projection direction of the data in the process of dimensionality reduction. In this paper, the idea of information entropy is introduced into the maximum Variance portfolio (MVU). In this paper, a nonlinear fault detection method based on KECA-MVU is proposed, which uses Rayleigh entropy to measure the effective degree of information retention after data transformation of kernel matrix obtained from MVU learning. According to the eigenvector corresponding to the first few terms of the maximum Rayleigh entropy as the projection direction of the data. Finally, the linear regression method is used to estimate the optimal projection matrix from the input data to the low-dimensional structure, and the on-line fault detection is realized by the projection matrix. Finally. This paper summarizes the main research results, and describes the difficulties and prospects of future research work.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP277

【參考文獻】

相關(guān)期刊論文 前3條

1 羅林;蘇宏業(yè);謝磊;班嵐;;一種改進的最大方差展開投影在非線性過程監(jiān)測中的應(yīng)用[J];江南大學(xué)學(xué)報(自然科學(xué)版);2014年06期

2 柴天佑;;生產(chǎn)制造全流程優(yōu)化控制對控制與優(yōu)化理論方法的挑戰(zhàn)[J];自動化學(xué)報;2009年06期

3 周東華;胡艷艷;;動態(tài)系統(tǒng)的故障診斷技術(shù)[J];自動化學(xué)報;2009年06期

相關(guān)博士學(xué)位論文 前6條

1 周哲;基于k近鄰的復(fù)雜工業(yè)過程故障診斷方法研究[D];浙江大學(xué);2016年

2 周樂;基于概率的工業(yè)過程數(shù)據(jù)建模與故障檢測[D];浙江大學(xué);2015年

3 文巧鈞;基于狀態(tài)空間模型的復(fù)雜動態(tài)過程監(jiān)測方法研究[D];浙江大學(xué);2015年

4 苗愛敏;數(shù)據(jù)局部時空結(jié)構(gòu)特征提取與故障檢測方法[D];浙江大學(xué);2014年

5 祝志博;融合聚類分析的故障檢測和分類研究[D];浙江大學(xué);2012年

6 邵紀東;非線性過程監(jiān)測中的數(shù)據(jù)降維及相關(guān)問題研究[D];浙江大學(xué);2010年

相關(guān)碩士學(xué)位論文 前1條

1 陳新如;基于線性動態(tài)系統(tǒng)模型的過程監(jiān)測研究[D];浙江大學(xué);2016年

，

本文編號：1469695