基于改进蚁群算法在面波频散曲线反演中的应用

Application of the Improved Ant Colony Algorithm in the Inversion of Rayleigh Wave Dispersion Curves

摘要:

为快速准确的反演得到近地表地层结构，将一种新颖而强大的非线性算法——蚁群算法引入到瑞雷波频散曲线领域，并对其进行相应的改进，改进蚁群算法的优点是运算效率快、精度高、算法简单、灵活易于实现，需要调节控制参数也较少。文中分别在无噪声\，含噪声以及实测数据进行反演测试，通过模型数据和实测数据表明，应用于瑞雷波反演中的改进蚁群算法在收敛速度与收敛精度之间能达到良好的平衡，所得解具有较高可信度。而且算法为促进所得解快速收敛到全局最优，在搜索中分全局搜索与局部搜索两个方式进行，能够有效地避免局部最优解产生。借助人工合成的瑞雷波数据以及真实观测数据，验证了改进蚁群算法在反演近地表剪切波速度时的有效性和通用性。此外，文中与遗传算法进行比较，得出改进蚁群算法具有高效性和高精度性的优点。

Abstract:

In this paper, we planned and improved the ant colony algorithm to quickly and accurately obtain the near-surface stratum structure. The ant colony algorithm is the simple algorithm, which has the advantages of fast operation efficiency and high precision. In this paper, the model data and the measured data indicated that the improved ant colony algorithm applied in Rayleigh wave inversion can achieve a better balance between the fast convergence rate and accuracy, and the solution has high credibility. This algorithm was divided into two ways, global search and local search, which can effectively avoid the generation of local optimal solution and thereby promoting the solution rapidly converge to the global optimum. Finally, the effectiveness and versatility of the improved ant colony algorithm in the inversion of near-surface shear wave velocity were verified using the artificial Rayleigh wave data and real observation data. Moreover, by comparing with the genetic algorithm, it was found that the improved ant colony algorithm has the advantages of high efficiency and accuracy.