A parallel shuffled complex evolution model calibrating algorithm to reduce computational time
A parallel shuffled complex evolution model calibrating algorithm to reduce computational time
复制标题
DOI：
--
复制
发表时间：
2007-12
期刊：
影响因子：
--
通讯作者：
N. Muttil;S. Liong;O. Nesterov
中科院分区：
文献类型：
--
作者：
N. Muttil;S. Liong;O. Nesterov
混洗复合进化算法（SCE-UA）是一种稳健、高效的搜索算法，已被广泛应用于径流模型的参数率定。尽管它的优越性，因为许多常用的径流模型有很大的模拟时间，模型校准算法的使用可能会变得不切实际，由于涉及的高计算时间。这将需要在校准算法中使用上级并行计算技术，目的是减少计算时间。本研究旨在将SCE-UA并行化以加速模型校准过程。在这项研究中，据观察，SCE-UA的并行化导致“线性加速”的模型校准过程中。这意味着，如果校准运行需要20小时才能完成1000个模型模拟，那么使用4台PC并行运行将需要5小时才能完成相同数量的模型模拟。此外，使用通过局域网连接的个人电脑并行化SCE-UA是一种简单和负担得起的替代方案，可以显著减少计算时间，而无需求助于昂贵的高端系统。这种计算时间的节省又促进了在校准过程期间对参数空间的显著更多的搜索。
The Shuffled Complex Evolution (SCE-UA) method has been widely applied for calibration of rainfall-runoff models and has been shown to be robust and efficient search algorithm. In spite of its superiority, since many commonly used rainfallrunoff models have large simulation times, the use of model calibrating algorithms may become impractical due to the high computational time involved. This would necessitate the use of superior parallel computing technologies in the calibrating algorithms, with the aim of reducing the computational times. This study aims to parallelize the SCE-UA to accelerate the model calibrating process. In this study, it is observed that the parallelization of SCE-UA leads to ‘linear speedups’ in the model calibrating process. This means that if a calibration run requires 20 hours to complete 1000 model simulations, then using 4 PCs in parallel would take 5 hours to complete the same number of model simulations. Moreover, parallelization of SCE-UA using PCs that are linked by a LAN is an easy and affordable alternative to achieve significant reduction of the computational time, without resorting to expensive high-end systems. Such savings in computational time in turn facilitates significantly more search of the parameter space during the calibration process.