基于角域信号重构及柔性聚焦的曲面多层结构超声检测方法

Curved multilayer structures are widely used in key equipments due to their advantage of large strength-to-weight. To guarantee the application security,it is very important to inspect the internal defects. As result of waveform excursion,overlapping and distortion in ultrasonic inspection for the curved multilayer structures, the ultrasonic signal features are difficult to extract ,and moreover, the resolutin of ultrasonic imaging is low. To solve the problems, an ultrasonic feature extraction method based on an angular domain signal reconstruction and a flexible focusing imaging theory will be studied in this project. Firstly, the propagation characteristics of ultrasonic in the curved multilayer structure will be investigated. The modular multi-Gaussian ultrasonic model is proposed to research the transformation law of echo signals in the angular domain.The signal reconstruction algorithm in the angular domain will be established by the cointegration theory. Then the echoes and defect waves will be separated and the ultrasonic characteristic parameters of the curved multilayer structures will be accuratly extracted. Secondly, an adaptive delay algorithm considering the difference of geometry and the speed of sound will be proposed according to the reconstructed characteristic parameters. Subsequently, the synthetic aperture optimization and the flexible planning methods for focal points will be researched. At the same time, the synthetic aperture focusing of a plane homogeneous structure will be developed to the flexible focusing of the curved multilayer structure, which will improve imaging resolution and reduce the computational complexity. This research is expected to provide theoretical guidance for explaining the transformation mechanism of ultrasonic signal and improving the ultrasonic inspecting for curved multilayer structures. The achievements have important theoretical significance and application value to improve the manufacturing quality and the safety in service of key equipments.

具有轻质及高强度优点的曲面多层结构被广泛应用到关键设备上，有效检出其内部缺陷是保障设备运行安全的前提。本项目针对曲面多层结构超声无损检测时波形偏移、混迭和畸变等造成信号特征提取困难及成像分辨率低的问题，拟开展基于角域信号重构的超声特征提取方法及柔性聚焦成像理论的研究，包括：（1）研究超声在曲面多层结构中的传播特性，构建模块化多元高斯超声模型，探求角域回波信号特征变换规律，基于协整理论提出角域信号重构算法，实现各层回波和缺陷波的分离，准确提取曲面多层结构超声信号特征参数。（2）采用重构后的特征向量，设计自适应聚焦延时算法，研究合成孔径优化及聚焦点柔性规划方法，将平面匀质结构合成孔径聚焦提升至曲面多层结构的柔性聚焦，提高成像分辨率并降低计算复杂度。预期成果为揭示曲面多层结构超声信号特征变换机理、改进复杂结构的超声检测方法提供理论依据，对提高关键设备制造质量与服役寿命具有重要理论意义及应用价值。

航空航天、兵器、电力、船舶、汽车等众多领域对曲面多层构件的无损检测有重要需求，本项目针对曲面多层结构超声无损检测时波形偏移、混迭和畸变等造成信号特征提取困难及成像分辨率低的问题，开展了基于角域信号重构及柔性聚焦的曲面多层结构超声检测方法研究。建立了曲面多层构件的超声传播模型，包括超声波传播声场、缺陷响应、系统测量函数等，研究了构件内部缺陷的回波特性及定量方法，获得了曲面构件缺陷定量表征曲线，实现了无需对比试块即可进行缺陷定量。通过对六自由度检测台运动学控制，实现了角域位姿精确控制和角域信号的采集，设计了基于特征优选的智能算法，准确分离出了各层回波与缺陷波。以水浸式多层构件检测为对象，考虑各层声速、路径变化所造成的渡越时间及位置偏移，提出了角域信号的重构算法，获取了准确的角域B扫成像，降低了斑点噪声的影响。同时为提高图像分辨率，改进了传统的平面匀质结构合成孔径聚焦算法，实现了多层结构合成孔径聚焦，且采用均方根速度模型提高了成像效率。在此基础上，提出了基于角域虚拟源模型的延时叠加算法，有效结合合成孔径聚焦成像与空间复合成像技术，仅需一次延时叠加计算即可获得角域超声合成孔径聚焦图像，减小了计算复杂度。本项目目前已累计发表论文18篇，其中SCI收录4篇，EI收录6篇。授权发明专利4项、实用新型专利2项、软件著作权1项。培养研究生10人，其中已毕业3人。获得湖南省科技进步二等奖1项。本项目的研究成果为完善复杂结构的超声检测方法提供了依据，对提高关键设备制造质量与服役寿命具有重要理论意义及应用价值。