免疫进化算法在复合极值分布参数估计中的应用

Application of Immune Evolutionary Algorithm to Parameter Estimate of Compound Extreme Model

参数估计是复合极值模型应用的难点之一,针对这一问题,利用免疫进化算法来估计复合极值模型的参数,给出在气候统计方面的两个应用实例。结果表明:免疫进化算法应用于复合极值模型参数求解是合理和可行的;相比于L-矩估计法及复合矩估计法,免疫进化算法所求参数拟合效果更好;免疫进化算法具有高效、普适的优点,在其他参数优化问题上也有广泛的应用前景。

Parameter optimization is becoming one of the difficulties for the application of compound extreme model. To solve this problem, immune evolutionary algorithm （IEA） is proposed to optimize the parameter of compound extreme model, which was demonstrated to be effective using two concrete examples. The results show that：It is reasonable and feasible to optimize the parameter of compound extreme model with the proposed IEA ;The fit model parameters calculated from IEA is better than that calculated from L-Moment and PWM ; With the advantages of highly efficient and general, IEA is potential to be used in solving other parameters optimization problems.