Lamb waves are characterized by frequency dispersion and mode conversion. The imaging results obtained with traditional ultrasonic array imaging methods suffer from artifacts, reducing the imaging accuracy. In addition, traditional methods are subjected to the Rayleigh criterion and cannot image multiple defects with spacing smaller than the resolution threshold of ultrasound array imaging. In this paper, an ultrasonic Lamb wave inverse scattering topological imaging method based on time-domain topological energy (TDTE) was proposed. In this method, the topological progressive process of the inverse scattering topological imaging method was transformed into solving the direct and the adjoint sound fields. Then, by inverting the time of the adjoint sound field, the direct and the adjoint sound field can focus adaptively at the defect, which overcomes the influence of Lamb wave mode conversion and dispersion on the imaging results, and improves the imaging accuracy and resolution. Finally, the direct and the adjoint sound field are fused, and the TDTE value was taken as the pixel value of imaging. Accordingly, the limitation of the Rayleigh criterion was broken; the resolution of imaging was improved; the multi asymmetric defect imaging with the interval was realized less than the resolution threshold of the ultrasonic array imaging. The experimental results show that the TDTE method for multiple asymmetric defects can achieve super-resolution than the total focus method (TFM) and time-reversal (TR) method and can accurately distinguish the asymmetric defects when the defect spacing is smaller than the resolution threshold of the ultrasonic array imaging.

兰姆波的特点是频率色散和模式转换。传统超声阵列成像方法获得的成像结果存在伪影，降低了成像精度。此外，传统方法受瑞利准则约束，无法对间距小于超声阵列成像分辨率阈值的多个缺陷进行成像。本文提出了一种基于时域拓扑能量(TDTE)的超声波兰姆波逆散射拓扑成像方法。该方法将逆散射拓扑成像方法的拓扑递进过程转化为求解直接和伴随声场。然后，通过将伴随声场的时间倒置，直接和伴随声场可以自适应地聚焦在缺陷处，克服了兰姆波模式转换和色散对成像结果的影响，提高了成像精度和分辨率。最后将直接声场和伴随声场融合，取TDTE值作为成像的像素值。因此，瑞利准则的限制被打破了；成像分辨率提高；实现了间隔小于超声阵列成像分辨率阈值的多非对称缺陷成像。实验结果表明，多非对称缺陷的TDTE方法比全聚焦法（TFM）和时间反转（TR）方法可以实现超分辨率，并且在缺陷间距小于分辨率阈值时可以准确地区分非对称缺陷超声阵列成像。