以定性模式為基礎之失誤診斷方法研究

陳俊彥

3.232.129.123，Hello！

Close
Search Quick Search Advanced Search Publication Search	Personal Service Member Login Searching History Use Bonus Card SYMSKAN
Browse Browse by University of Graduate Thesis Browse by Pinyin of Graduate Thesis Browse by Conference Proceedings Topic Browse by Conference Proceedings Publisher Browse by Pinyin of Conference Proceedings Browse by eBook Category Browse by eBook Topic Browse by eBook Publisher
Customer Services Authorize Airiti Recommend a Journal

Search Symbol (Half-width)	Description of Search Symbols
Space	"AND" indicates the intertwining of key terms used in a search
Double Quotation Marks ("") ( " " )	Double quotation marks indicate the beginning and end of a phrase, and the search will only include terms that appear in the same order of those within the quotations. Example: "image process" ： " image process "
?	Indicates a variable letter. Entering two ? will indicate two variable letters, and so on. Example: "Appl?", search results will yield apple, apply… e , appl y … ( (often used to English word searches) )
*	Indicates an unlimited number of variable letters to follow, from 1~n. Example: Enter "appl*", search results will yield apple, apples, apply, applied, application…(often used in English word searches) e , appl es , appl y , appl ied , appl ication … ( (often used to English word searches) )
AND、OR、NOT	Boolean logic combinations of key words is a skill used to expand or refine search parameters. (1) AND (1) AND: Refines search parameters (2) OR (2) OR: Expands search parameters (3) NOT: Excludes irrelevant parameters

DOI stands for Digital Object Identifier （ D igital O bject I dentifier ） ,
and is the unique identifier for objects on the internet. It can be used to create persistent link and to cite articles.

Using DOI as a persistent link

To create a persistent link, add「http://dx.doi.org/」「 http://dx.doi.org/ 」 before a DOI.
For instance, if the DOI of an article is 10.5297/ser.1201.002 , you can link persistently to the article by entering the following link in your browser: http://dx.doi.org/ 10.5297/ser.1201.002 。
The DOI link will always direct you to the most updated article page no matter how the publisher changes the document's position, avoiding errors when engaging in important research.

Cite a document with DOI

When citing references, you should also cite the DOI if the article has one. If your citation guideline does not include DOIs, you may cite the DOI link.

DOIs allow accurate citations, improve academic contents connections, and allow users to gain better experience across different platforms. Currently, there are more than 70 million DOIs registered for academic contents. If you want to understand more about DOI, please visit airiti DOI Registration （ doi.airiti.com ）。

Data Source

College of Engineering > Department of Chemical Engineering
Engineering > Chemical Industry

Bibliography Management Tool

Related Links

Report an Issue

Search in Library

There is no full-text available at the moment. Airiti would like to request author's authorization. I would like to provide authorization.

以定性模式為基礎之失誤診斷方法研究

Studies on Qualitative Model-Based Fault Diagnosis Methods

陳俊彥 , Ph.D　　Advisor：張珏庭　　

英文

失誤診斷；有號有向圖；感測器的配置；自動機；模糊系統； fault diagnosis ； signed directed graph ； sensor placement ； automata ； fuzzy

Abstract │ Reference (115) │ Altmetrics

Abstract 〈TOP〉

由於製程的尺度及複雜性急遽成長，失誤發生所造成的經濟利益及安全危害勢必更加嚴重，因此失誤診斷在化學程序中已成為一個不可缺少的任務。本論文探討兩個在失誤診斷上極為重要的議題。第一個議題針對誤源預測來做研究，第二個議題則是探討感測器的配置策略。對於第一個問題本文提出一種兩階段的策略，此兩階段分別為（1）線下準備階段（2）線上實作階段。在線下階段（1）中，本文採用以有號有向圖為基礎的程序，此程序可定性的預測當一失誤發生或多個失誤發生時所有可能的失誤徵兆及失誤演化順序並且依據此預測產生相對應的模糊推論系統。我們可將此程序大致分為下面幾個步驟:建構診斷器，預測失誤徵兆及失誤演化順序，生成語言，建構模糊推論系統。
 建構診斷器:首先利用有號有向圖模組來預測每一個系統狀態可能後續發生的事件並由初始的正常狀態遞迴地利用此預測來產生所有系統狀態。其中事件的預測不僅包括不同變數之間的傳播 (空間序列傳播) 亦包括同一變數間不同大小等級的傳播 (時間序列傳播)。此外我們可以轉換所有狀態及狀態間的事件為一組節點及邊以便構成狀態轉移圖或者稱之為自動機。最後若將自動機上線上無法測量的事件離去便可以獲得對應的診斷器。
  預測失誤徵兆及失誤演化順序:根據產生的診斷器我們可以觀察到所有的狀態既符合所有可能的失誤徵兆並且由初始狀態沿著事件所代表的邊直到穩態所形成的所有路徑正符合每一個可能的失誤演化序。
  生成語言:每一個由上一步驟產生的路徑皆包含了一連串的事件序列，我們可以將此事件序列視為一個字串，因此所有的路徑對應的所有字串即可構成一組正規語言。
  建構模糊推論系統:最後所有的狀態及字串可能被編碼成為兩種不同的“若則”法則。此兩種不同的法則分別代表了靜態失誤徵兆及動態失誤傳播順序的對應法則。根據這些法則我們可以建構出兩層式的模糊推論系統，其中每一層分別對應於上述的兩種“若則”法則。一旦模糊系統建構完成在線上階段（2）中只需將線上測量數據傳送到此系統即能給出每一失誤可能發生的指標。最後以五個實例來論證此法的可行性。
    此外針對第二個問題本文結合了有號有向圖以及失誤演化順序來改進文獻中所提出的策略。此方法依據失誤演化順序來增加額外的資訊以便提供更好的感測器選擇。此資訊主要依據“兩個誤源若含有不同演化順序既可能用此不同順序的測量值來辨別此兩個不同的源”，因而這些測量值便可能形成感測器的候選位置。為了將此額外訊息併入文獻所提出的方法，本文亦提出了轉換此訊息到雙分矩陣的程序並且修正了文獻所提出了貪婪演算法以適用於擴張型的雙分矩陣。除此之外本文也嘗試以整數線性規畫的方式來解決此一問題並以三個實例來論證此兩種解法的可行性以及差異性。

Parallel Abstract 〈TOP〉

Due to the increase of chemical process scale and complexity, unpredicted fault will influence the economic interest and the process safety even more than it did before. Therefore, on-line fault diagnosis system should be considered as an indispensable tool in modern chemical process. Two fault diagnosis problems are studied in this thesis. First problem concerns about fault isolation, and the second problem concerns about sensor placement strategy. For the first problem, a method with two stages is proposed, i.e.,(1) the off-line preparation stage and (2) the on-line implementation stage. On stage (1), a SDG-based reasoning procedure is presented in this thesis to qualitatively predict all possible symptom patterns and also their evolution sequences caused by one or more fault propagating in any given process system as well as constructing the corresponding fuzzy inference system. This procedure consists of the following steps: generation of diagnoser, generation of symptom patterns and evolution sequences, generation of language and construction of fuzzy inference system.

generation of diagnoser: The SDG model is used to predict all possible propagated events under a system state and recursively suggest all possible states derived from the initial normal state. Notice the predicted propagated events not only involve the propagation between different variables (spatial propagation) but also the propagation between different magnitudes of a same variable (temporal propagation). Consequently, these states and predicted events between states can be represented as nodes and edges of state-transition digraph, i.e., automata. In addition, the automata can also be converted into a diagnoser by
deleting all unobservable measurements.

generation of symptom patterns and evolution sequences: From the diagnoser, the corresponding possible symptom patterns as well as evolution sequence which represents the path from initial state to the final state caused by the fault are generated in this step.

generation of language: Every path constructed by the previous step contains one consecutive events sequence, which can be identified as a word or a sting. Later, strings corresponding to all the paths can be combined together to construct a formal language.

construction of fuzzy inference system: Finally, all states in diagnoser and strings can be encoded into two different kinds of IF-THEN rules. These two kinds of IF-THEN rules represent the static fault symptoms(candidate patterns) and the dynamic fault symptoms (evolution sequences)respectively. From these rules a comprehensive fuzzy inference system can be constructed for the online diagnosis purpose. The comprehensive fuzzy inference system consists of two layer fuzzy structure. The first layer represents the static fault diagnosis and the second layer represents the dynamic fault diagnosis. On stage (2), online measurements are fed into the fuzzy inference system constructed on stage (1) to show the possible outcome caused by every fault occurred. The feasibility of the proposed method is demonstrated here by using five different examples.

For the second problem, a SDG-based method is proposed to design the sensor network on the basis of qualitatively predicted fault evolution sequences. This method is able to improve the original strategy proposed in the literature by taking the fault evolution sequences into consideration as additional information to assist in selecting sensor. The information is based on "If two fault origins contain different evolution sequences, the measurements of those different sequences can be used to distinguish those two different fault origins" principle and able to determine the placement of the sensors. To combine this additional information to the existing method proposed by literature, this thesis provide procedures to convert the information into bipartite matrix and revise the existing greedy algorithm using expanded bipartite matrix. Furthermore, to achieve a maximum level of resolution and, at the same time, to ensure observability, the corresponding design problems are also formulated as integer linear programs. The feasibility of the proposed method along with two different strategies, i.e., integer programs and greedy algorithm, is demonstrated with three different examples.

Reference ( 115 ) 〈TOP〉

Ali, Y., Narasimhan, S., 1993. Sensor network for maximizing reliability of linear process. AIChE J. 39(5), 820.
連結：
Ali, Y., Narasimhan, S., 1995. Redundant sensor network for maximizing reliability of linear process. AIChE J. 41(10), 820.
連結：
Ali, Y., Narasimhan, S., 1996. Sensor network for maximizing reliability of bilinear process. AIChE J. 42(9), 2563.
連結：
Bagajewicz, M., Carbrera, E., 2002. A new milp formulation for instrumentation network design and upgrade. AIChE J. 48(10), 2271.
連結：
Bagajewicz, M., Fuxman, A., 2004. Instrumentation network design and upgrade for process monitoring and fault detection. AIChE J. 50(8), 1870.
連結：

Altmetrics

〈TOP〉

E-mail ：

When an article is available to download, a notice will be sent to your mailbox address.

E-mail ：