基于神经网络和遗传算法的泥沙模型研发及应用  

Research and Development of Sediment Model Software Based on Artificial Neural Network and Genetic Algorithms

作　　者：陈雄波 王俊昀 龚钰婷 程冀 CHEN Xiongbo;WANG Junyun;GONG Yuting;CHENG Ji(Yellow River Engineering Consulting Co.,Ltd.,Zhengzhou 450003,China)

机构地区：[1]黄河勘测规划设计研究院有限公司,河南郑州450003

出　　处：《人民黄河》2020年第12期18-22,共5页Yellow River

基　　金：国家重点研发计划项目(2017YFC0405506)。

摘　　要：针对黄河流域的泥沙问题,根据BP网络原理和训练流程、遗传算法原理,优化BP网络参数的加速遗传算法的迭代步骤,编制了计算程序;根据数学模型计算及管理的特点,设计完成了软件功能表和模型功能模块;在研发的软件中检验了计算程序模块的功能,经测试是正确的;将计算模块无缝嵌入到模型软件中,并通过方案管理、结果查询等辅助模块实现了计算过程的通用化、可视化。该模型软件的组成模块在黄河口流路运用方案比选、多泥沙河流断面概化中得到应用,经分析成果比较合理。Aiming at the sediment problem in the Yellow River basin,according to the principle of BP network and training flow,genetic algorithm and the iteration steps of accelerated genetic algorithm to optimize the parameters of BP network,and the calculation program was compiled.According to the characteristics of mathematical model calculation and management,the software function table and model function module were designed and completed.The calculation program was tested and verified.The calculation module was seamlessly embedded in the model software and the computational process was universal and visualized through auxiliary modules such as scheme management and result query.The component module of the model software had been applied in the comparison and selection of the operation schemes of the Yellow River estuary and the generalization of the sedimentladen river section.The analysis results were reasonable.

关 键 词：BP神经网络 遗传算法 优化模型 计算过程 通用化 可视化 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TV142[自动化与计算机技术—控制科学与工程]