基于金屬磁記憶效應的井下套管預測故障研究

1.9萬字
我自己原創(chuàng)的畢業(yè)論文，僅在本站獨家提交，大家放心使用

摘要： 大慶油田的井下套管損壞（套損）嚴重，近80 %的套損是由于地質因素和工程因素造成的，這些因素使得套管受到非均勻載荷的作用而產(chǎn)生應力集中或者微觀裂紋，進而導致套管彎曲、變形或錯斷。傳統(tǒng)的射線，漏磁檢測等方法只能檢測已經(jīng)出現(xiàn)裂紋的套管，無法預測由應力集中導致疲勞的早期故障，使得套管壽命預測和預報受到了限制。而金屬磁記憶檢測技術可以有效判斷出套管應力集中區(qū)，做到預測和預報早期診斷；利用基于磁記憶效應的檢測系統(tǒng)可以快速、準確地檢測在役套管的應力集中區(qū)所產(chǎn)生的磁記憶信號，從而可快速、準確地判斷套管的應力集中區(qū)和故障狀況。本論文圍繞金屬磁記憶信號的采集和特征量的提取進行研究，準確確定應力集中區(qū)，消除外界噪聲的影響。論文主要研究內容包括：
　　1. 介紹了金屬磁記憶檢測技術的發(fā)展歷程，闡述了金屬磁記憶早期診斷鐵磁構件疲勞失效的重要意義。
   2. 金屬磁記憶信號是地磁場在套管應力集中區(qū)域的“記憶”表現(xiàn)，信號很微弱，且易受到外界噪聲污染，從而影響實際應用中對套管損傷程度的準確評估。對井下數(shù)據(jù)采用數(shù)字平滑技術預處理，并利用傅里葉變換進行頻譜分析，從而獲得磁記憶信號的主要能量集中區(qū)。在此基礎上，引入切比雪夫低通濾波器對信號進行消噪處理，利用梯度極大值特征準確判斷套管的應力集中區(qū)。
  3．為了進一步提取磁記憶信號的奇異特征量，從時——頻角度對磁記憶信號進行分析。采用一維離散小波變換和自適應閾值方法對信號進行多尺度分解，將信號的低頻部分和高頻部分全部顯示出來。通過與之前的頻域分析對比，可知低頻部分最為接近原始信號，這樣避免了其他頻段對磁記憶信號的干擾。
  4. 提出了新的組合特征量診斷方法，通過BP神經(jīng)網(wǎng)絡實現(xiàn)了磁記憶信號特征量和安全等級的非線性關系，保證可靠地判斷套管應力集中危險區(qū)域。通過實驗驗證所提方法的有效性和可行性。
   關鍵詞：套管；金屬磁記憶；故障；預測；應力集中；小波分析

Casing failure predicting based on magnetic memory
Abstract：The damage of the damaging oilfield of downwhole casing is sever. Nearly 80%of the damage of the casing is due to geological factors and engineering factors which lead to  the casing led to tube bending, deformation or dislocation because of the role of non-uniform load and stress concentration or micro-cracks.traditional method for example ray detection and the detection of the magnetic flux leakage casing cracks can only detect the cracks which has occurred, but not to predict early failure caused by the Stress concentration .The life of the casing and the prediction of the casing are restricted because of the situation.however The metal magnetic memory technology can effectively determine the casing stress concentration area, to predict and forecast the result of early diagnosis .Based on the use of magnetic memory effect detection  the technology can quickly and accurately detect the magnetic memory signals of stress concentration due to the stress concentration area, which which can quickly and accurately determine the casing stress concentration area and fault conditions.main contend of the paper  is about  the collection and feature extraction of metal magnetic memory signal to study which accurately determine the stress concentration area, eliminating the effects of outside noise.The main research contents include:
1 Introduces the development of metal magnetic memory testing technology, describes the importance of early diagnosis of ferromagnetic metal magnetic memory component fatigue failure.
2. Magnetic memory signal is the performance of the memory of the stress concentration area owning to the geomagnetic field. the signal is very weak and vulnerable to external noise pollution, thus affecting the accurate assessment of the extent of damage to the casing during the practical application.Downhole data preprocessing using digital smoothing techniques, and the use of Fourier transform spectral analysis can obtain the stress concentration aera.On this basis, the introduction of Chebyshev low-pass filter the signal denoising using gradient maxima characterized accurately determine stress concentration cannula.
3.In order to extract the singular characteristic quantity of magnetic memory signal, we analysis the signal from the angle of  the time --frerquency. One-dimensional discrete wavelet transform and adaptive threshold method for multi-scale decomposition of the signal, the low-frequency and high-frequency portion of the signal part of the whole show. By contrast with the previous frequency domain analysis, we can see that some of the signal will be most close to the original low-frequency signal, thus avoiding interference with other bands magnetic memory signal.
4. Proposed a new combination of features volume diagnostics, BP neural network to achieve a non-linear relationship between the magnetic memory signal features and security levels can ensure casing stress concentration reliably determine the danger zone. the experiment proposes the method which can verify the effectiveness and feasibility.
Keyword：Casing; Metal magnetic memory; Fult; Forecasting; Stress Concentration; Wavelet analysis

目 錄

摘要 I
Abstract II
目 錄 1
第-章  緒 論 3
1.1引言 3
1.2研究背景和意義 4
1.3國內外研究現(xiàn)況 4
1.4論文研究內容 8
第二章 金屬磁記憶檢測套管信號處理方法 9
2.1 磁記憶信號采集 9
2.2  數(shù)據(jù)實驗 9
2.2.1金屬套管加力實驗設計 9
2.2.1.1 套管加壓設施模型設計 10
2.2.1.2  套管加力實驗裝置 13
2.3原始采樣信號分析 14
2.4磁記憶信號的頻域分析與濾波 16
2.4.1磁記憶信號的頻域結果分析 17
2.4.2磁記憶信號的切比雪夫濾波 18
2.5套管磁記憶信號的小波分析 21
2.5.1套管磁記憶信號的強制去噪 21
2.5.2套管磁記憶信號的閾值去噪 27
第三章  套管應力集中的磁記憶非線性定量反演 33
3.1  金屬磁記憶的非線性定量反演 34
3.2  金屬磁記憶信號的特征量提取 34
3.3 神經(jīng)網(wǎng)絡 39
3.3.1  BP神經(jīng)網(wǎng)絡算法 39
3.4  實驗結果 45
結論 48
參考文獻 49