基于优化并行模态超声导波的管道腐蚀量化方法研究

Corrosion in pipeline is the critical factor that could have a destructive effect on the reliability of pipeline system and its service life. As one of important problems involved in pipeline inspection, the quantitative evaluation of pipeline corrosion has great challenges. Ultrasonic guided waves can enable long-rang, comprehensive and efficient inspection. However, due to its irregular profile, the reflection generated by corrosion is very complicated, resulting in the absence of quantitative measurement of pipeline corrosion. This project aims to propose a new method for guided wave-based inspection with the optimized parallel modes used, so that the information of corrosion embedded in reflection can be improved given the different sensitivities of the parallel modes to the various features of corrosion; to define the new concept of reflection coefficient vector (RCV) to present the full characters of reflections at different corrosion features, and to study the interaction between guided waves and pipeline corrosion by investigating the relationship among wave modes, corrosion features and RCV; to built the strategy for optimizing parallel modes concerning their types and parameters through a multiple-object optimization procedure with the RCV as evaluation indicator; to study how to apply sparse decomposition method and deep study algorithm for the quantitative evaluation to corrosion features. It is expected to build a new scheme that can enable an efficient, comprehensive evaluation of pipeline corrosion in quantitative way through these research works. This proposed project could provide a scientific support for the safety evaluation and maintenance of pipeline system, preventing any possible incidence caused by corrosion so that the whole system could run reliably. Therefore, this project has high values in both the academic research and practical applications.

管道腐蚀是影响管网可靠性及使用寿命的关键因素，腐蚀的量化评估是管道检测的重要问题和难点。超声导波技术可提供大范围、全面高效的检测，但由于腐蚀形态不规则，产生的反射回波复杂，目前尚不能实现管道腐蚀的精细测量。本项目针对此问题提出一种基于优化并行模态的超声导波检测新方法，它利用不同模态对腐蚀各特征具有不同敏感性增强反射回波中的有效信息；定义反射系数向量，以表征腐蚀各特征方向上回波的量化特性；通过研究导波模态、腐蚀特征和反射系数多维分量之间的关系分析导波与管道腐蚀的交互作用；以反射系数向量为优化目标，对模态组合及其时空参数进行多目标优化，并建立并行模态优化策略；进而采用稀疏分解技术结合深度学习算法对管道腐蚀特征进行定量评估，最终形成一种新的管道腐蚀量化的解决方案。本研究有望为改善管道安全评估与维护提供科学对策，从而预防事故发生，确保管网可靠运行，因此不仅具有较高学术价值，且具有重要的应用前景。

管道运输业是国民经济的命脉，管道在使用过程中可能出现腐蚀、疲劳、蠕变，材质劣变和冲刷磨损等多种破坏形式，而其中管道腐蚀已成为影响管道系统可靠性及使用寿命的关键因素。腐蚀检测评估及相关管道安全问题一直是工业界和学术界普遍关注的重要课题。超声导波技术超可有效弥补传统无损检测方法的诸多不足，但因管道腐蚀形态不规则，导波检测产生的腐蚀回波复杂，目前尚不能实现管道腐蚀的精细测量。本项目针对超声导波的复杂管道缺陷检测及评估问题，对超声导波与各类管道缺陷的交互过程,及复杂缺陷特征的定量化检测策略进行了深入研究与分析。主要包括：在对管道缺陷特性研究的基础上，建立形成了管道导波及复杂缺陷仿真模型，提出了基于多量化评估指标综合分析实现管道导波缺陷检测仿真系统的优化方案，保证了超声导波建模仿真的准确和快速；设计了多种类型的导波信号收发装置，研究了模态控制等问题，设计搭建了管道缺陷检测的超声导波管道远程检测实验系统，开展了大量实验研究。针对不同管道缺陷与导波交互的仿真和实验研究表明，复杂缺陷的边界特征对回波有较大影响，缺陷回波的波形由缺陷边界特征决定，缺陷的径向深度与周向宽度主要影响缺陷回波的反射能量，因此提出了缺陷特征边界的概念并提出了基于特征边界反射的管道缺陷量化策略。对管道点蚀散射特性的研究发现，腐蚀与单模态导波会产生强烈的模态转换现象，造成缺陷回波严重衰减且增加了信号的复杂性。基于以上结果，进一步提出了基于宽频共振的缺陷特征量化评估方法，利用多模态对不同缺陷特征的共振响应实现缺陷特征的确定与提取。本项目可为解决管道复杂缺陷特征辨识与结构健康评价等问题提供新思路和新方法，具有重要的科学意义和工程价值。