混沌差分算法在确定河流水质模型参数中的应用

Application of chaotic difference evolution algorithm in determining model parameter of river water quality

将混沌寻优思想引入到差分优化算法形成混沌差分算法,并将其应用于确定河流水质模型参数的函数优化问题。数值实验结果表明:应用混沌差分算法求解此参数问题无论是在精度还是时间上都优于差分优化算法。它将混沌寻优的遍历性和随机性思想引入到差分优化算法中,在每次差分进化寻得的最优位置附近进行混沌细搜索,并配合特殊的迭代终止准则进行寻优。其明显缩短了混沌搜索计算时间和克服了差分优化算法后期早熟的缺陷,提高模型求解的收敛速度和精度。

The paper introduced the idea of chaos optimization into differential optimization algorithms so as to form chaotic differential algorithm, and applied it to the function optimization problems of model pa- rameter of river water quality. The results show that the chaotic differential algorithm is better in solving the parameter problem than the differential algorithm on accuracy and time. It brought the chaos optimiza- tion ideas of ergodieity and randomness into differential optimization algorithm, and carried out chaotic detail search near the optimal location after running difference method in each iteration, cooperated with special termination criterion to look for optimization. The method significantly shortened the calculation time of chaotic search , overcame the precocious defects of difference algorithm in the later period and improved the convergence speed and accuracy of the model.