張士華 黃松嶺 孫永泰 史永晉 王宏安

摘 要：漏磁檢測(cè)是在管道內(nèi)檢測(cè)中應(yīng)用最廣泛的一種無(wú)損檢測(cè)技術(shù)，檢測(cè)數(shù)據(jù)量化與分析是氣難點(diǎn)。在技術(shù)方面針對(duì)課題重點(diǎn)研究的關(guān)鍵技術(shù)開(kāi)展了一系列研究，提出了油氣管道漏磁檢測(cè)數(shù)據(jù)的分類和量化方法，并基于此研發(fā)出一套漏磁檢測(cè)數(shù)據(jù)分析軟件。漏磁檢測(cè)中缺陷量化困難的原因在于缺陷的形態(tài)對(duì)漏磁場(chǎng)的形態(tài)有復(fù)雜的非線性的影響，繼而影響對(duì)漏磁信號(hào)的定量解釋，因此，根據(jù)缺陷的開(kāi)口形狀將缺陷進(jìn)行分類，對(duì)于實(shí)現(xiàn)將其準(zhǔn)確量化是十分必要的。再者，由于實(shí)際檢測(cè)條件的限制，往往只能通過(guò)空間離散的漏磁感應(yīng)強(qiáng)度信號(hào)的一維分量推算缺陷的三維形態(tài)，這本身不適合使用精確的數(shù)學(xué)或者統(tǒng)計(jì)模型加以描述。使用神經(jīng)網(wǎng)絡(luò)對(duì)缺陷進(jìn)行量化，是漏磁檢測(cè)缺陷量化領(lǐng)域近20年來(lái)的一個(gè)研究熱點(diǎn)。根據(jù)課題研究?jī)?nèi)容以及檢測(cè)器設(shè)計(jì)指標(biāo)，提出了一種基于改進(jìn)徑向基函數(shù)網(wǎng)絡(luò)的量化算法，它以缺陷漏磁場(chǎng)信號(hào)的特征量為輸入，輸出向量為缺陷的三維外形參數(shù)。徑向基函數(shù)網(wǎng)絡(luò)是一種局部最佳逼近網(wǎng)絡(luò)，但漏磁檢測(cè)中漏磁感應(yīng)強(qiáng)度信號(hào)與缺陷外形之間強(qiáng)烈的非線性關(guān)系，往往更要求所選用的網(wǎng)絡(luò)能夠識(shí)別兩者間的內(nèi)在聯(lián)系，并使得面對(duì)新的數(shù)據(jù)時(shí)仍有合理的量化結(jié)果。為此，對(duì)徑向基函數(shù)網(wǎng)絡(luò)做出基于泛化能力優(yōu)化的改進(jìn)，提出新的評(píng)價(jià)函數(shù)，并采用能夠迅速適應(yīng)新樣本的在線學(xué)習(xí)算法，實(shí)驗(yàn)驗(yàn)證表明，的確能大幅提高網(wǎng)絡(luò)的泛化能力。在實(shí)際工程檢測(cè)管道中，多缺陷聚集會(huì)明顯影響漏磁場(chǎng)的形態(tài)，軸向槽缺陷漏磁場(chǎng)與兩個(gè)坑狀缺陷信號(hào)波形極為相似，緩變?nèi)毕萋┐艌?chǎng)信號(hào)變化趨勢(shì)較小，這對(duì)定量漏磁檢測(cè)的實(shí)用化是不容忽視的問(wèn)題。討論了不同類型缺陷漏磁場(chǎng)形態(tài)和強(qiáng)度的影響，并測(cè)試了量化神經(jīng)網(wǎng)絡(luò)對(duì)缺陷間隔變化的適應(yīng)能力。研究以分類和量化算法為核心，研發(fā)一套漏磁檢測(cè)數(shù)據(jù)分析系統(tǒng)。該系統(tǒng)配合內(nèi)檢測(cè)器已項(xiàng)目中投入測(cè)試，對(duì)牽拉實(shí)驗(yàn)數(shù)據(jù)分析的結(jié)果驗(yàn)證了所提出算法的確具有優(yōu)秀的量化性能。

關(guān)鍵詞：漏磁檢測(cè) 缺陷分類 缺陷量化 多缺陷聚集 數(shù)據(jù)分析系統(tǒng)

Abstract：The magnetic flux leakage（MFL） is the most generalized method for in-pipe inspection. A method of classification and quantification of defects in MFL inspection is proposed， and a data analysis system is developed based on this method. The pattern of magnetic flux leakage has a complex non-linear relationship with the shape of defects， which makes it a difficult problem to make quantitative analysis to the magnetic flux leaked.Furthermore， in reality testing conditions， usually only the component in one direction is detected for quantification. Such problems do not adapt to accurate mathematical models. Utilizing neural network as a quantification method has become a focus in MFL inspection during the last 20 years. A method of quantification based on modified radial base function neural network （RBFNN） is proposed. RBFNN promises locally optimal approximation， but the non-linear relationship between magnetic flux pattern and the defect shape requires a strong capability to recognize their inner connection， to better deal with generalized samples.Anon-line trainingalgorithm to determine the number of nodes in hidden layer is proposed， and new merit function based on optimized generalization is employed to train the central vectors and widths. Both of them， verified by experiments， can greatly enhanced the generalized capability of RBFNN. Corrosions usually appear as multi-defect assemblies in pipelines. The relationship between magnetic flux leakage and the pattern of multi-defect assembly is discussed. And different neural network models are employed to solve the inverse problem for multi-defect assembly. Based on the methods stated above， a data analysis expert system is developed. This system works coordinating with in-line inspector and is tested in a submerged pipeline in-service testing project. Results prove that the modified methods gives accurate predicts to a wide range of defects.

Key Words：Magnetic Flux Leakage Inspection；Classification of Defects；Quantification of Defects；Multi-defect Assembly；Data Analysis System

閱讀全文鏈接（需實(shí)名注冊(cè)）：http：//www.nstrs.cn/xiangxiBG.aspx？id=65390&flag=1