改进蜂群算法及其在面波频散曲线反演中的应用

Improvement of Artificial Bee Colony and its application in Rayleigh wave inversion

应用改进蜂群算法反演面波频散曲线以获得近地表横波速度剖面.蜂群算法属于群智能算法中的一种,灵感来源于蜜蜂群体特定的觅食行为,在该算法的基础上结合粒子群算法中的全局最优解引导思想,同时引入遗传算法中交叉运算操作,即采用基于交叉操作的全局人工蜂群算法对面波频散曲线进行反演研究.改进蜂群算法在继承传统算法精于探索特性的同时,针对其疏于开发的缺陷着重加强了算法对全局的探索能力.使用理论和实测瑞雷波数据,本文研究了改进蜂群算法在推导近地表横波速度分布的有效性和适用性.在反演中,目标函数的收敛性好,改进算法在迭代的过程中能够快速收敛到全局最优;模型参数的概率分布高,即在寻找到全局最优解的同时,能够确保解中每个参数同时达到最优,保证了反演的结果可靠度,使其能有效地应用于瑞雷波频散曲线的反演和解释中.

The improved artificial bee colony algorithm is applied to get surface wave phase velocities.The ABC algorithm,one of swarm intelligence-based algorithms,was inspired from the particular intelligent foraging behavior of honeybee swarm in nature.On the basis of the ABC algorithm,this paper combines with the idea of global optimal solution in particle swarm optimization algorithm and the cross-operation strategy in genetic algorithm.The improved algorithm on the basis of inheriting traditional ABC algorithm,can also improve the exploration ability.Using theoretical and measured data,this paper increased the effectiveness and applicability of the improved ABC algorithm in deducing an S-wave velocity profile in near-surface applications.In inversion,the objective function can rapidly converge to the global optimization solution.The wide probabilitydistribution of model parameters means that it is able to find the global optimal solution and at the same time to ensure each parameter in the solution to achieve the best.It guaranteed the improved ABC algorithm is suitable for the inversion and interpretation of Rayleigh wave dispersion curve.