超声回波信号的连续极性一致法分离谱处理

Consecutive polarity coincidence based SSP to ultrasonic reflected signal

给出了一种基于连续极性一致恢复算法的超声回波信号分离谱处理方法,该方法不需要有关输入信号频带的先验知识,不必预先设置滤波器组的中心频率等参数,且能自动锁定缺陷回波和实时调节阈值以优化处理结果。数值仿真结果均表明,使用该方法可以提高信噪比,有效提取缺陷回波,且性能优于其他常规分离谱方法。

A method of split-spectrum processing （SSP） for ultrasonic reflected signal was proposed, which employs consecutive polarity coincidence as the recovering algorithm. This method need neither to know the priori knowledge of frequency band of input signal nor to set the center frequency of filter banks in advance. Meanwhile, it can adaptively lock flaw echo and adjust threshold to optimize the results of processing in real time. The numerical simulation results manifested that, by using this method, the ratio of signal to noise can be improved and the flaw echo can be effectively extracted with better performance than other conventional SSP methods.