粒子群算法在主震断层面参数估计中的应用*

(防灾科技学院,河北三河 065201)

粒子群算法; 参数估计; 非线性优化模型; 余震; 主震断层面

Application of Particle Swarm Optimization to the Estimation of mainshock Fault Plane Parameters
WANG Fu-chang,WAN Yong-ge,HU Shun-tian

(Institute of Disaster Prevention Science and Technology,Sanhe 065201,Hebei,China)

Particle Swarm Optimization; parameter estimation; Nonlinear Optimization Model; aftershock; fault plane of mainshock

备注

提出了利用粒子群全局优化算法,由余震震源参数确定主震断层面的方法,从而克服了常见的Gauss-Newton法对初值的敏感性。同时,讨论了l1和l2范数准则对参数估计的影响。将该方法用于2003年7月21日云南大姚6.2级地震断层面参数估计,取得了较好的结果。

A new method to determine the fault plane of mainshock by using Particle Swarm Global Optimization Algorithms was proposed,which overcomes the sensitivity of Gauss-Newton Method to initial value.The effect of l1- and l2- criteria on parameter estimation was also discussed.This method was used to evaluate the fault plane parameters of the Dayao,Yunnan,earthquake on July 21,2003,and the results show that the method is valid.