To analyze the influence of various possible errors in the actual ground-penetrating radar data processing on the imaging effect of the reverse time migration algorithm, numerical simulation tests are used to analyze the effect of the velocity errors in model and noises in data. For velocity error, models with different velocities and abnormal body depths, and simplified velocity models are used to test the subsequent imaging effect. Results show that with the increase of the velocity error and the depth of the abnormal body, the deviation of the position of the abnormal body also increases. For the layered structure with slight differences in velocity between each layer, imaging using a simplified uniform half-space model can still effectively distinguish the anomalies. Imaging tests are also performed on data with different noise levels. The results show that the reverse time migration (RTM) algorithm is effective on noise suppression. Even if the signal-to-noise ratio of the valid signal in the received data is around ?15 dB, the RTM algorithm can still image the corresponding structure.

为了分析探地雷达数据处理过程中各种可能的误差对逆时偏移算法成像效果的影响, 通过数值模拟测试, 对模型存在速度误差和数据存在噪声的情况进行分析。对于模型存在速度误差的情况, 选用不同模型速度、不同异常体深度以及简化速度模型的情况进行成像效果测试。结果表明, 随着速度误差加大, 异常体位置偏离也变大, 且异常体深度越大, 位置偏离越大。对于各层速度相差不大的层状结构, 使用简化的均匀半空间模型进行成像, 依然可以较好地分辨地下的异常体。对于存在噪声的情况, 选用加入不同噪声水平的数据进行成像测试, 结果表明逆时偏移(RTM)算法对噪声有较好的压制效果, 即使接收的数据中有效信号的信噪比在?15 dB左右, 该算法仍然能将对应的结构成像出来。