论文部分内容阅读
粘弹性参数变得越来越重要,其反演算法也逐渐成为众多研究者的研究热点。而遗传算法是一种随机、自适应、启发式的算法, 具有很好的鲁棒性和全局收敛性, 本文基于VSP直达波方程,引入了遗传算法来进行粘弹参数反演, 首先将频率域直达波方程表示为复速度的函数,然后通过遗传算法反演出复速度。而复速度和品质因子又是复速度的函数,从而便可很容易的得出。但若直接反演复速度, 反演参数太多, 不容易实现, 所以又将复速度表示成参数C0和C∞的函数,以减少反演参数数量。最后给出了理论模型实验,以证明该算法的有效性。
Viscoelastic parameters become more and more important, and the inversion algorithm has gradually become the research hotspot of many researchers. Genetic algorithm is a stochastic, adaptive and heuristic algorithm with good robustness and global convergence. Based on the VSP direct wave equation, genetic algorithm is introduced to perform viscoelastic parameter inversion. Firstly, the frequency The domain direct wave equation is expressed as a function of complex velocities, and then the complex velocities are retrieved by genetic algorithms. The complex speed and quality factor is a function of complex velocity, which can be easily derived. However, if the inversion speed is directly inverted and too many inversion parameters are not easy to implement, the complex velocity is expressed as a function of parameters C0 and C∞ to reduce the number of inversion parameters. Finally, a theoretical model experiment is given to prove the effectiveness of the algorithm.