【摘要】 针对提高混凝土结构超声无损检测分辨率的需要,提出了两种处理方法。在对低频B扫描数据进行合成孔径聚焦处理时,由于低频探测信号时间宽度大,波包会被拉伸,发生波包变形现象,为此,引入波包分解技术来解决这个问题。波包分解技术实质是用一些模型参数去描述检测信号,可以避免对回波信号所有采样点进行处理,不仅解决了波包变形问题,而且可以简化计算。对于实验中商用超声传感器引起的余振问题,提出一种数字滤波的补偿方法,通过水浸方式采集数据建立传感器系统补偿模型,用于实际检测信号中振铃的抑制,可提高成像的时空分辨率。数值仿真和实验结果证明,采用这两种方法后内嵌目标定位理准确,成像分辨率明显提高。更多还原

【Abstract】 Two signal processing approaches are presented to improve imaging resolution in ultrasonic NDT of con- crete structures.When low-frequency B-scan data are processed by Synthetic Aperture Focusing Technique (SAFT), wavepacket will be further elongated because low-frequency detection signal covers a relative long period.Wavepacket Decomposition Technique (WDT) is introduced here to solve this problem.This method uses a few parameters to describe the original signal so that it can avoid the processing of the whole time history of signals.In practical mea- surement,ringing of commercial ultrasonic transducers also affects the discrimination of the detected signal,a method based on digital filtering is proposed to build the compensation model of transducer system.By eliminate the ringings with the filter,the spatiotemporal resolution of ultrasonic imaging is improved.Numerical simulation and experimental results have proved the efficiency of the methods.Imaging resolution is improved obviously and the embedded object in a test specimen is located accurately.更多还原