本文選題：腐蝕預(yù)警　切入點(diǎn)：數(shù)據(jù)挖掘　出處：《西安電子科技大學(xué)》2015年碩士論文　論文類型：學(xué)位論文

【摘要】：油管柱腐蝕是影響石油開(kāi)采時(shí)間成本以及物資成本的一個(gè)重要因素。油管柱腐蝕會(huì)降低油管柱的工作壽命,降低石油開(kāi)采效率。如何準(zhǔn)確預(yù)測(cè)工作環(huán)境中油管柱腐蝕速率,更換與維護(hù)油管柱從而降低石油開(kāi)采的時(shí)間與物資成本,是石油開(kāi)采行業(yè)的一個(gè)重要問(wèn)題。如果能從大量的油管柱腐蝕試驗(yàn)數(shù)據(jù)中發(fā)現(xiàn)出油管柱的腐蝕規(guī)律,可以減少大量人力物力資源的消耗。由于目前在油管柱腐蝕預(yù)測(cè)領(lǐng)域上對(duì)腐蝕數(shù)據(jù)的處理大多數(shù)使用傳統(tǒng)的方法,分析維度低,大量腐蝕數(shù)據(jù)的信息沒(méi)有得到有效的利用。本文引入數(shù)據(jù)挖掘技術(shù),利用這一技術(shù)提取腐蝕數(shù)據(jù)的有效信息。本文首先對(duì)油管柱腐蝕預(yù)警問(wèn)題進(jìn)行分析,總結(jié)油管柱腐蝕預(yù)警特征,然后對(duì)油管柱腐蝕預(yù)警需要解決的問(wèn)題進(jìn)行分析,總結(jié)出解決油管柱腐蝕預(yù)警問(wèn)題至少需要滿足要求,在此基礎(chǔ)上,提出運(yùn)用數(shù)據(jù)挖掘技術(shù)解決油管柱腐蝕預(yù)警問(wèn)題的方法。方法的基本思想是通過(guò)數(shù)據(jù)挖掘方法分析油管柱腐蝕試驗(yàn)數(shù)據(jù),得出腐蝕規(guī)律,在使用中把油管柱工作環(huán)境參數(shù)傳遞給腐蝕規(guī)律,計(jì)算腐蝕速率,及時(shí)反饋給管理人員。本文分別將樸素貝葉斯算法,貝葉斯網(wǎng)絡(luò)算法等貝葉斯算法,C4.5算法,隨機(jī)森林算法等決策樹(shù)算法,以及人工神經(jīng)網(wǎng)絡(luò)算法應(yīng)用于油管柱腐蝕規(guī)律發(fā)現(xiàn),預(yù)測(cè)和預(yù)測(cè)模型的創(chuàng)建,并對(duì)其效果進(jìn)行分析。分析結(jié)果顯示,基于人工神經(jīng)網(wǎng)絡(luò)算法的數(shù)據(jù)挖掘方法更適合解決油管柱腐蝕預(yù)警問(wèn)題。本文對(duì)數(shù)據(jù)挖掘技術(shù)中的人工神經(jīng)網(wǎng)絡(luò)算法進(jìn)行了詳細(xì)的研究,分析其預(yù)測(cè)模型建立以及訓(xùn)練過(guò)程,得出人工神經(jīng)網(wǎng)絡(luò)對(duì)任意函數(shù)有很好的逼近能力,通過(guò)調(diào)整優(yōu)化參數(shù)可以很好地應(yīng)用在油管柱腐蝕試驗(yàn)數(shù)據(jù)挖掘中。本文將人工神經(jīng)網(wǎng)絡(luò)算法應(yīng)用到油管柱腐蝕預(yù)警數(shù)據(jù)挖掘程序中。油管柱腐蝕預(yù)警數(shù)據(jù)挖掘程序是油管柱腐蝕預(yù)警系統(tǒng)的預(yù)測(cè)模塊,通過(guò)對(duì)油管柱腐蝕數(shù)據(jù)進(jìn)行分析并處理,建立預(yù)測(cè)模型并對(duì)需要預(yù)測(cè)的環(huán)境條件進(jìn)行油管柱腐蝕速率預(yù)測(cè)。本文基于數(shù)據(jù)挖掘技術(shù)與Weka工具使用人工神經(jīng)網(wǎng)絡(luò)模型對(duì)油管柱腐蝕預(yù)測(cè)數(shù)據(jù)挖掘程序進(jìn)行開(kāi)發(fā),利用Tomcat,Maven和Nexus工具,對(duì)其Web程序進(jìn)行開(kāi)發(fā)。為驗(yàn)證研究成果的可行性與有效性,本文設(shè)計(jì)并開(kāi)發(fā)了一個(gè)基于數(shù)據(jù)挖掘的油管柱腐蝕預(yù)測(cè)程序,介紹了系統(tǒng)的架構(gòu)設(shè)計(jì)、流程設(shè)計(jì)、程序設(shè)計(jì)和關(guān)鍵代碼實(shí)現(xiàn)。應(yīng)用效果表明,本文提出的數(shù)據(jù)挖掘技術(shù)Web技術(shù)解決油管柱腐蝕預(yù)警問(wèn)題的方法是可行有效的。
[Abstract]:The corrosion of tubing string is an important factor affecting the time cost and material cost of oil production. The corrosion of tubing string will reduce the working life of tubing string and reduce the efficiency of oil production. How to accurately predict the corrosion rate of tubing string in working environment? Replacing and maintaining tubing string so as to reduce the time and material cost of oil production is an important problem in petroleum production industry. If the corrosion law of tubing string can be found from a large number of corrosion test data of tubing string, It can reduce the consumption of a lot of human and material resources. Because the traditional method is used to deal with the corrosion data in the field of pipeline corrosion prediction, the dimension of analysis is low. The information of a large number of corrosion data has not been effectively utilized. In this paper, data mining technology is introduced to extract the effective information of corrosion data. This paper summarizes the characteristics of tubing string corrosion warning, then analyzes the problems that need to be solved, and concludes that the problem of tubing string corrosion warning should at least meet the requirements. This paper puts forward a method of using data mining technology to solve the problem of pipe string corrosion warning. The basic idea of the method is to analyze the corrosion test data of tubing string by data mining method, and to obtain the corrosion law. In use, the operating environment parameters of tubing string are transferred to corrosion law, corrosion rate is calculated, and feedback is given to managers in time. In this paper, Bayesian algorithms such as naive Bayes algorithm, Bayesian network algorithm, etc. The decision tree algorithm such as stochastic forest algorithm and artificial neural network algorithm are applied to the discovery, prediction and prediction of the corrosion law of tubing string, and the effect of the model is analyzed. The data mining method based on artificial neural network algorithm is more suitable to solve the problem of pipeline corrosion warning. In this paper, the artificial neural network algorithm in data mining technology is studied in detail, and its prediction model establishment and training process are analyzed. It is concluded that the artificial neural network has good approximation ability to arbitrary functions. In this paper, the artificial neural network algorithm is applied to the data mining program of tubing string corrosion early warning data mining. Sequence is the prediction module of tubing string corrosion warning system. By analyzing and processing the corrosion data of tubing string, Based on the data mining technology and Weka tool, the pipeline corrosion prediction data mining program is developed by using artificial neural network model. In order to verify the feasibility and validity of the research results, a pipeline corrosion prediction program based on data mining is designed and developed in this paper. The architecture design and flow design of the system are introduced. Program design and key code implementation. The application results show that the method of data mining technology Web proposed in this paper is feasible and effective to solve the problem of pipe string corrosion warning.
【學(xué)位授予單位】：西安電子科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：TE983;TP311.13

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 張海濤;黨問(wèn)清;;關(guān)于對(duì)油田采油管柱技術(shù)應(yīng)用探討[J];中國(guó)石油和化工標(biāo)準(zhǔn)與質(zhì)量;2012年13期

2 陳旭峰;;探討油田采油管柱技術(shù)應(yīng)用[J];化工管理;2012年S1期

3 邵魁歧;;對(duì)油田采油管柱技術(shù)應(yīng)用探究[J];化工管理;2013年20期

4 蔡亞西,施太和,王幼金;連續(xù)油管柱振動(dòng)分析[J];西南石油學(xué)院學(xué)報(bào);1998年01期

5 劉迪軍,張廣清,仇偉德,馮啟寧;油管柱內(nèi)壓和軸力傳感器的實(shí)驗(yàn)研究[J];石油大學(xué)學(xué)報(bào)(自然科學(xué)版);2001年02期

6 樁棟,劉煥偉,顏?lái)?鄧玉華,孫玉華;油管防落裝置的研制與應(yīng)用[J];石油礦場(chǎng)機(jī)械;2003年06期

7 王禾丁,谷開(kāi)昭,朱愛(ài)萍;各種油管錨定方法淺析[J];石油礦場(chǎng)機(jī)械;2003年06期

8 崔旭明,何富君,譚云,楊曉翔,羅敏,張永弘;油管的疲勞強(qiáng)度分析[J];應(yīng)用力學(xué)學(xué)報(bào);2003年01期

9 范興亮,何春容,何世明,李榮;用計(jì)算機(jī)校核油管柱強(qiáng)度的方法[J];鉆采工藝;2005年02期

10 謝守平;;可增加下入深度的變徑連續(xù)油管柱[J];石油機(jī)械;2005年12期

相關(guān)會(huì)議論文 前2條

1 高德利;;石油管柱力學(xué)與工程研究進(jìn)展[A];科技、工程與經(jīng)濟(jì)社會(huì)協(xié)調(diào)發(fā)展——中國(guó)科協(xié)第五屆青年學(xué)術(shù)年會(huì)論文集[C];2004年

2 李海峰;;石油管柱下井遇阻碰撞的數(shù)值模擬及系統(tǒng)集成開(kāi)發(fā)[A];中國(guó)工程物理研究院科技年報(bào)(2008年版)[C];2009年

相關(guān)博士學(xué)位論文 前1條

1 蹇清平;基于漏磁檢測(cè)的油管缺陷量化研究[D];西南石油大學(xué);2015年

相關(guān)碩士學(xué)位論文 前10條

1 張佩;油井生產(chǎn)桿管系統(tǒng)優(yōu)化研究[D];長(zhǎng)江大學(xué);2015年

2 袁健威;基于數(shù)據(jù)挖掘的油管柱腐蝕預(yù)測(cè)預(yù)警問(wèn)題研究與系統(tǒng)開(kāi)發(fā)[D];西安電子科技大學(xué);2015年

3 楊競(jìng)哲;油管斷裂失效的分析[D];吉林大學(xué);2010年

4 張學(xué)東;抽油機(jī)井油管判廢技術(shù)研究[D];大慶石油學(xué)院;2009年

5 姚紅燕;油水井油管的腐蝕與防護(hù)措施研究[D];中國(guó)石油大學(xué);2007年

6 王光普;抽油機(jī)井油管的斷裂強(qiáng)度分析[D];吉林大學(xué);2011年

7 何磊;油管內(nèi)涂抗磨新材料及涂覆工藝的研究與應(yīng)用[D];中國(guó)石油大學(xué);2007年

8 陳朝暉;特殊螺紋油管入井質(zhì)量控制技術(shù)研究[D];西安石油大學(xué);2012年

9 童斌;玻璃鋼油管在油水井應(yīng)用技術(shù)研究[D];中國(guó)石油大學(xué);2011年

10 王鵬;高壓高產(chǎn)氣井試氣用油管柱抗擠毀性能研究[D];西安石油大學(xué);2013年

，

本文編號(hào)：1646871