基于改进智能优化算法的投影寻踪模型在洪水评估中的应用研究

Study on application of improved intelligence optimization algorithms-based projection pursuit model to flood assessment

传统的洪水评估方法存在着评估等级离散和结果不易分辨等不足,如何更准确高效地解决洪水评估问题已成为研究领域的热点之一。以南京站的历史洪水及四川省历史洪水灾情为例,在改进智能优化算法的基础上,引入了基于智能优化算法的投影寻踪模型,并探讨该模型在洪水分类和洪灾等级评价中的应用。结果表明,人工蜂群和混合蛙跳这类新型智能优化算法具有简单、鲁棒、全局寻优和易于实现等特点,与广泛应用于水文界的SCE-UA、文献中的加速遗传等现代启发式算法相比,具有寻优速度更快、能力更强的优势,可为洪水分类和洪灾等级评价等相近领域研究提供新途径。

By taking the historical floods recorded in Nanjing Hydrological Station and the historical flood disasters occurred in Sichuan Province as the study cases, the improved intelligence optimization algorithms-based projection pursuit model is introduced herein on the basis of improving intelligence optimization algorithms, and then the applications of the model to the flood classification and the evaluation of flood disaster grade are discussed. The result shows that the new artificial intelligence optimization algorithms, such as artificial bee colony algorithm and shuffled frog leaping algorithm, have the features of simplicity, robustness, global optimization and easy realization ; which have the advantages of quicker optimizing speed and stronger optimizing capacity, if compared with the heuristic algorithms, such as SCE - UA used in the field of hydrology and those accelerating genetic algorithms given in the literatures concerned, and then can be provide a new way for the studies made in the similar fields of flood classification, evaluation of flood disaster grade, etc.