探地雷达数值模拟及其在隧道检测中的应用

Numerical Simulation of Ground Penetrating Radar and Its Application in Tunnel Detection

利用时域有限差分(FDTD)方法,对隧道工程检测中常见检测目标进行数值模拟,分析脱空、单层钢筋、双层钢筋等目标体的探地雷达反射信号图像特征,并对隧道检测中的雷达信号图像采用遗传算法进行反演,引入目标相关系数用于反演结果全局寻优,实现对多次反射的压制,降低浅部钢筋反射信号对识别深部目标体的影响。通过对数值模拟结果进行反演,证明该方法理论上的有效性,并结合工程实践中的应用效果,说明该方法的优点与局限性。

The common detecting targets of tunnel engineering detection are simulated by means of the finite-difference time-domain （FDTD） method. The features of the ground penetrating radar （GPR） signal reflected by void, single layer and double-layer rebar and other objects are analyzed, the genetic algorithm （GA） is applied in the inversion of GPR tunnel detection image, and the target correlation coefficient is introduced for the global optimization of inversion results, so as to suppress the multiple reflection and reduce the impact of shallow rebar reflection on the recognition of deeper target. The effectiveness of the method is demonstrated by the inversion of the numerical simulation results, and the application results of the engineering practices can justify the advantages and limitations of this method.