PTA裝置溶劑脫水塔軟測量建模及靈敏度分析

化工論文編號:HG219  論文字數:19945,頁數:40

目  錄
中文摘要 I
英文摘要 II
目錄 III
1 緒論 1
1.1軟測量模型及其方法............................................ 1
1.1.1 機理建模.................................................2
1.1.2 回歸分析.................................................2
1.1.3 狀態估計.................................................2
1.1.4 基于模式識別.............................................3
1.1.5 神經網絡.................................................3
1.1.6 模糊系統.................................................4
1.2 存在的問題....................................................4
1.3 模型校正......................................................5
1.4 化工過程的操作優化............................................5
1.5 PTA裝置溶劑脫水塔的現狀......................................6
2. 軟測量建模方法——回歸分析 8
2.1引言..........................................................8
2.2多元線性回歸(MLR).............................................9
2.2.1多元線性回歸模型推導.....................................9
2.2.2實例驗證................................................10
3. 軟測量建模方法一人工神經網絡.......................................13
3.1　引言.........................................................13
3.2　神經網絡的特點和及應用.......................................14
3.3　多層前傳網模型...............................................15
3.3.1　網絡拓撲結構............................................15
3.3.2　目標函數和訓練算法......................................16
3.4　實例驗證.....................................................19
4.回歸分析與ANN結合的混和建模........................................23
4.1　引言..........................................................23
4.2  PCA－BP神經網絡的軟測量建模.................................23
4.2.1　輸入變量的初選...........................................23
4.2.2　隱層節點數的確定.........................................24
4.2.3  PCA－BP神經網絡模型結構................................24
4.2.4  BP神經網絡的LM優化算法................................24
4.3實例驗證.......................................................25
5.實驗結果與討論.......................................................28
5.1 實驗結果......................................................28
5.2 討論..........................................................28
6. 總結與展望..........................................................31
6.1 本文工作小結..................................................31
 6.2 今后的研究發展方向............................................32
致謝 ...................................................................33
參考文獻 ...............................................................34

摘 要：本文主要以PTA裝置溶劑脫水塔為研究對象，運用統計回歸方法和人工神經網絡技術，建立溶劑脫水塔出口產品組成的PCA-BP神經網絡軟測量模型。本文完成的工作主要有：
1)本文在系統分析了基于PCA-BP的建模原理后，對人工神經網絡模型的關鍵參數即隱層節點數的確定作了較為深入的探討，總結出了比較有效的取值方法。并在PTA裝置溶劑脫水塔產品組成的軟測量建模中得到應用，取得了較為滿意的結果。
2)研究了基于統計回歸的溶劑脫水塔塔頂塔底產品組成的軟測量建模方法。并選用多元線性回歸(MLR)方法對溶劑脫水塔產品組成進行建模，對所建模型進行了驗證，其結果表明模型的精度達到了預期的要求。
3)研究了基于BP算法的人工神經網絡對溶劑脫水塔產品組成進行建模的方法，并用現場實測數據對所建模型進行驗證測試。
本文最后，對統計回歸和神經網絡技術在過程軟測量建模的應用進行了展望，并提出了今后建模過程中需要進一步深入研究的方向。

關鍵詞：溶劑脫水塔；多元線性回歸；軟測量；主成分回歸；人工神經元網絡

Abstract: This paper is a study of PTA solvent dehydration tower. The author used a statistical regression method and the technology of artificial neural network, and built a soft-sensing model of the PCA-BP neural network which made up the solvent dehydration tower products for export. In this paper included the following aspects: firstly, the author analyzed the main parameter of the artificial neural network based on the principles of PCA-BP modeling, and summed up a more effective method. And this method has been applied to achieve a satisfactory result. Secondly, the author added up the soft-sensing approaches to both the top and bottom products of solvent dehydration tower. Based on these approaches, the author chose a multiple linear regression (MLR) method in modeling the solvent dehydration tower products, and also inspected those models. The result showed the accuracy of those models has achieved the expected standard. Thirdly, the author studied the method based on the BP artificial neural network in modeling the solvent dehydration tower products, and tested the models through the data on the scene. At last, the author took a long view on statistical regression method and the technology of artificial neural network used in soft-sensing modeling, and then pointed out some ideas for further work.

Keywords: Solvent Dehydration Tower; Multiple Linear Regression (MLR); Soft-sensor; Principal Component Regression(PCR); Artificial Neural Network.