Weibull分布参数的粒子群算法估计

Parameters Estimation for Weibull Distribution with Particle Swarm Optimization

基于广西涠洲岛海洋监测站3个方向的年极值波高观测资料,在假设其服从Weibull分布的基础上,运用最小二乘法,矩估计法和最速下降法对Weibull分布的参数进行估计,同时引入粒子群算法确定Weibull分布的3个参数,对文中4种方法得出的拟合结果及运算效率进行比较分析,说明了粒子群算法在估计极值分布参数中的优势。

Based on the annual maximum wave heights in the three directions measured at Guangxi Weizhoudao Observation Station, this paper estimates the three unknown parameters of Weibull distribution with least-square method, moment method and the steepest descent algorithm. Meanwhile the parti cle swarm optimization is also introduced to determine the parameters of Weibull distribution. Through the comparison of the above 4 methods, the particle swarm optimization shows its superiority in the calcu lation of statistical parameters for Weibull distribution.