本文選題：高斯過程回歸 + 模型更新��； 參考：《江南大學(xué)》2017年碩士論文

【摘要】：實(shí)際工業(yè)過程多具有非線性、時(shí)變性及不確定性等特點(diǎn),而傳統(tǒng)離線軟測量模型無法對此類工業(yè)過程的狀態(tài)參數(shù)進(jìn)行實(shí)時(shí)跟蹤。針對上述問題,通常對傳統(tǒng)離線軟測量方法進(jìn)行自適應(yīng)改進(jìn),并根據(jù)實(shí)時(shí)數(shù)據(jù)對在線軟測量模型參數(shù)及數(shù)據(jù)庫進(jìn)行一定的預(yù)處理和更新,以確保所建軟測量模型具備跟蹤過程動態(tài)特征及抗干擾的能力,最終進(jìn)一步提高模型精度以及性能。為了實(shí)現(xiàn)對過程主導(dǎo)變量進(jìn)行有效預(yù)測及控制,本文首先采用高斯過程回歸(Gaussian Process Regression,GPR)算法對實(shí)際工業(yè)過程進(jìn)行學(xué)習(xí)并得到相應(yīng)軟測量模型,隨后提出動態(tài)模型更新及奇異點(diǎn)檢測補(bǔ)償算法對其進(jìn)行動態(tài)校正和預(yù)處理,最終通過實(shí)驗(yàn)仿真論證了本文所提算法的有效性。論文的主要研究內(nèi)容如下所示:(1)高斯過程回歸算法實(shí)際應(yīng)用研究。首先對高斯過程回歸算法原理進(jìn)行簡要解析。隨后利用此回歸算法對青霉素發(fā)酵過程進(jìn)行學(xué)習(xí)并建立相應(yīng)軟測量模型。通過與傳統(tǒng)最小二乘支持向量機(jī)(Least Squares Support Vector Machine,LSSVM)的仿真對比,表明所建高斯過程回歸模型具有更好的預(yù)測性能。(2)進(jìn)一步考慮工業(yè)過程的時(shí)變特征,提出一種基于動態(tài)模型更新的GPR在線軟測量方法。該方法首先對訓(xùn)練樣本利用GPR方法進(jìn)行離線建模,得到預(yù)測輸出及預(yù)測誤差;然后對所得預(yù)測誤差進(jìn)行分析,當(dāng)誤差均值大于某一預(yù)設(shè)閾值時(shí)對GPR模型進(jìn)行整體更新:同時(shí)更新其協(xié)方差矩陣和協(xié)方差函數(shù)的參數(shù);否則,只對GPR模型進(jìn)行局部更新:只更新其協(xié)方差矩陣。接著利用誤差高斯混合模型(Error Gaussian Mixture Model,EGMM)對更新后的GPR模型預(yù)測輸出進(jìn)行誤差補(bǔ)償從而得到最優(yōu)的預(yù)測結(jié)果。最終由實(shí)際工業(yè)污水處理過程的實(shí)例仿真驗(yàn)證了所提方法的有效性。(3)針對軟測量方法在實(shí)際應(yīng)用中查詢樣本可能出現(xiàn)奇異點(diǎn)這一問題,提出一種帶奇異點(diǎn)檢測補(bǔ)償?shù)腉PR在線軟測量方法。該方法首先對訓(xùn)練樣本利用GPR方法進(jìn)行建模,得到軟測量模型;然后對新來查詢樣本采用改進(jìn)拉依達(dá)準(zhǔn)則進(jìn)行奇異點(diǎn)檢測,當(dāng)新來查詢樣本被確定為奇異點(diǎn)時(shí),利用輔助模型對奇異點(diǎn)進(jìn)行修補(bǔ),然后再利用軟測量模型對修補(bǔ)后查詢樣本點(diǎn)進(jìn)行預(yù)測;否則,直接使用軟測量模型對新來查詢樣本點(diǎn)進(jìn)行預(yù)測。最終通過實(shí)際硫回收過程數(shù)據(jù)的實(shí)驗(yàn)仿真驗(yàn)證了所提方法的有效性。
[Abstract]:The actual industrial processes are usually nonlinear, time-varying and uncertain, but the traditional off-line soft sensor model can not track the state parameters of such industrial processes in real time. In order to solve the above problems, the traditional off-line soft sensing methods are usually improved adaptively, and the parameters of the online soft-sensing model and the database are preprocessed and updated according to the real-time data. In order to ensure the dynamic characteristics of tracking process and the ability of anti-jamming, the model can improve the precision and performance of the model. In order to effectively predict and control the process leading variables, this paper first uses Gao Si process regression Gaussian Process algorithm to study the actual industrial process and obtains the corresponding soft sensor model. Then a dynamic model updating and singular point detection compensation algorithm is proposed to dynamically correct and preprocess it. Finally, the effectiveness of the proposed algorithm is demonstrated by experimental simulation. The main contents of this paper are as follows: 1) Gao Si process regression algorithm. Firstly, the principle of Gao Si process regression algorithm is analyzed briefly. Then the regression algorithm was used to study the penicillin fermentation process and the corresponding soft sensor model was established. Compared with the traditional least square support vector machine (LSSVM), it is shown that the proposed Gao Si process regression model has better predictive performance and further considers the time-varying characteristics of industrial processes. An online soft sensor method for GPR based on dynamic model updating is proposed. In this method, the training sample is modeled off-line by GPR method, and the prediction output and prediction error are obtained, and then the prediction error is analyzed. When the mean error is greater than a preset threshold, the GPR model is updated as a whole: the covariance matrix and the parameters of the covariance function are updated at the same time; otherwise, the GPR model is only locally updated: only its covariance matrix is updated. Then the error Gaussian Mixture model EGMMM is used to compensate the error of the updated GPR model to get the best prediction result. Finally, the effectiveness of the proposed method is verified by a practical example of industrial wastewater treatment. 3) aiming at the problem that the sample may appear singularity in the application of soft sensor, This paper presents an online soft sensing method for GPR with singularity detection compensation. In this method, the training sample is modeled by GPR method, and the soft sensor model is obtained, and then the new query sample is detected by the improved Laida criterion, when the new query sample is determined as the singularity point. The singular points are repaired by the auxiliary model, and then the sample points are predicted by the soft sensor model. Otherwise, the new query sample points are predicted directly by the soft sensor model. Finally, the effectiveness of the proposed method is verified by the experimental simulation of the actual sulfur recovery process data.
【學(xué)位授予單位】：江南大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TQ018

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 王開榮;高佩婷;;建立在DY法上的兩類混合共軛梯度法[J];山東大學(xué)學(xué)報(bào)(理學(xué)版);2016年06期

2 孫茂偉;楊慧中;;基于改進(jìn)仿射傳播聚類的多模型軟測量建模及應(yīng)用[J];南京理工大學(xué)學(xué)報(bào);2016年02期

3 熊偉麗;張偉;徐保國;;一種基于EGMM的高斯過程回歸軟測量建模[J];信息與控制;2016年01期

4 于霜;劉國海;梅從立;程錦翔;;生物發(fā)酵過程中VIP優(yōu)化神經(jīng)網(wǎng)絡(luò)逆系統(tǒng)的軟測量方法[J];南京理工大學(xué)學(xué)報(bào);2015年04期

5 張偉;熊偉麗;徐保國;;基于實(shí)時(shí)學(xué)習(xí)的高斯過程回歸多模型融合建模[J];信息與控制;2015年04期

6 薛明晨;熊偉麗;徐保國;;基于局部加權(quán)偏最小二乘的在線多模型建模[J];計(jì)算機(jī)應(yīng)用研究;2015年10期

7 熊偉麗;姚樂;徐保國;;基于EM算法的青霉素發(fā)酵過程多階段融合建模[J];化工學(xué)報(bào);2014年12期

8 王振雷;唐苦;王昕;;一種基于D-S和ARIMA的多模型軟測量方法[J];控制與決策;2014年07期

9 王鑫;李紅麗;;苯酚含量預(yù)測的高斯過程回歸模型[J];自動化儀表;2014年05期

10 楊慧中;董陶;陶洪峰;;基于改進(jìn)K-means聚類算法的組合模型建模[J];控制工程;2013年02期

，

本文編號：1807894