隧道地震预报中基于遗传算法的波动方程反演

首先建立一维波动方程反演的数学模型，而后引入交错网格有限差分正演模拟，最后将遗传算法和交错网格差分正演结合，实现波动方程遗传算法的反演；针对数值模拟合成的单道记录，应用波动方程遗传算法反演分析各参数对反演结果的影响；采用工程实例验证反演的预报效果，比较前后两次预报相同里程段的波速值。结果显示，S波反演结果具有更好的一致性。

Genetic Algorithms Inversion of Wave Equation in Tunnel Seismic Prediction

Firstly,the mathematical model of one-dimensional wave equation is built, then staggered grid finite difference forward modeling is introduced, and finally the staggered grid difference forward and genetic algorithms are combined to achieve the inversion of wave equation. We apply the genetic algorithm inversion of wave equation to study the numerical synthesis of single-channel simulated recording, then analyze the impact of various parameters on the inversion results. Finally,we examine engineering examples to verify the effects of the wave equation prediction genetic algorithm inversion, including both forward and backward comparison prediction the same velocity value times segment mileage. S-wave inversion has better consistency, due to the S-wave energy and its ability to tolerate interference in the same mileage acquired segments close to the S-wave.