王庭辉,顾正华,周统,等. 涟漪算法及其在马斯京根模型参数优化中的应用[J]. 水利水运工程学报,2024.. doi: 10.12170/20221011004
引用本文: 王庭辉,顾正华,周统,等. 涟漪算法及其在马斯京根模型参数优化中的应用[J]. 水利水运工程学报,2024.. doi: 10.12170/20221011004
(WANG Tinghui, GU Zhenghua, ZHOU Tong, et al. The ripple algorithm and its application in optimizing the parameters of the Muskingum model[J]. Hydro-Science and Engineering, 2024(in Chinese)). doi: 10.12170/20221011004
Citation: (WANG Tinghui, GU Zhenghua, ZHOU Tong, et al. The ripple algorithm and its application in optimizing the parameters of the Muskingum model[J]. Hydro-Science and Engineering, 2024(in Chinese)). doi: 10.12170/20221011004

涟漪算法及其在马斯京根模型参数优化中的应用

The ripple algorithm and its application in optimizing the parameters of the Muskingum model

  • 摘要: 基于自然界的涟漪现象,提出一种新的启发式智能优化算法——涟漪算法(ripple algorithm, RA)。该算法模仿涟漪的结构,搜索过程采用由中心点出发的三层涟漪随机取点搜索,搜索中心点群以当前适应度与全局最优适应度的比较信息为基础;采用涟漪半径函数与收缩函数对涟漪的展开进行控制,相对独立地向解空间最优点收敛。通过测试函数将涟漪算法与粒子群算法、标准遗传算法、引力搜索算法进行对比测试。结果表明:涟漪算法在较低维度与运用广泛的启发式算法相比有很强的竞争力;在马斯京根模型参数优化应用中,涟漪算法寻优能力强、精度高,有很强的实用性。对涟漪半径函数、收缩函数及其他影响涟漪算法优化过程的参数进行讨论,提出了几种改进涟漪算法的思路与方向。

     

    Abstract: This paper introduces a new heuristic intelligent optimization algorithm called the ripple algorithm (RA), inspired by the ripple phenomenon in nature. The algorithm mimics the structure of ripples and utilizes a three-layer ripple random point search starting from the central point. The search center group is determined based on the comparative information between the current fitness and the global optimal fitness. The expansion and contraction of ripples are controlled by the ripple radius function and contraction function, respectively, allowing the algorithm to converge towards the best point in the solution space independently. Through comparisons with other commonly used heuristic algorithms such as particle swarm optimization algorithm, standard genetic algorithm, and gravity search algorithm using test functions, the results demonstrate that the ripple algorithm exhibits strong competitiveness in lower dimensions. Moreover, when applied to parameter optimization of the Muskingum model, the ripple algorithm demonstrates excellent optimization ability, high accuracy, and practicality. The paper also discusses the ripple radius function, contraction function, and other parameters that impact the optimization process of the ripple algorithm, and proposes several ideas and directions for further improving the algorithm.

     

/

返回文章
返回