利用粒子群优化算法快速、稳定反演瑞雷波频散曲线

反演瑞雷波频散曲线能有效地获取横波速度和地层厚度，但基于局部线性化的瑞雷波频散曲线反演方法很难适应反演目标函数的非线性、多参数、多极值的特点。为此，提出并测试了一种新的基于全局优化策略的粒子群优化（PSO）算法的瑞雷波频散曲线反演方法。首先反演了三个理论模型的无噪声和含噪声数据，验证了PSO对瑞雷波数据反演的有效性与稳定性；然后将PSO与模拟退火法（SA）进行对比，说明PSO相对于SA具有全局收敛性强、收敛速度快、求解精度高的特点；最后，反演了来自美国怀俄明地区的实测数据，检验了PSO对瑞雷波数据反演的适用性。理论模型试算和实测资料分析表明，PSO可以用于瑞雷波频散曲线的定量解释。

Fast and stable Rayleigh-wave dispersion-curve inversion based on particle swarm optimization

Rayleigh-wave dispersion-curve inversion can effectively obtain shear wave velocity and formation thickness.However,Rayleigh-wave dispersion-curve inversion based on local linearization cannot adapt to inversion objective function characteristics such as non-linear,multi-parameters,and multi-extremums.To overcome this issue,we propose a new Rayleigh-wave dispersion-curve inversion based on a particle swarm optimization (PSO) algorithm for global optimization.We first invert synthetic data with noise and without noise of three theoretical models,and verify the effectiveness and stability of the PSO inversion of Rayleigh wave data.Then we compare PSO with simulated annealing (SA),and find that PSO has faster convergence and higher accuracy than SA.Finally,we apply this method into field seismic data from Wyoming in the United States to test its applicability.Theoretical and real data tests show that the proposed method can be used for the quantitative interpretation.