基于遗传优化神经网络的市政管网水质模型研究  

Water quality model of municipal network based on genetic optimization neural network

作　　者：任彬[1] 周荣敏[1] 

机构地区：[1]郑州大学水利与环境学院,河南郑州450002

出　　处：《供水技术》2010年第3期31-34,共4页Water Technology

摘　　要：将遗传算法的全局搜索能力和BP神经网络的局部学习能力有机结合,得到一种快速高效的建立供水管网余氯的水质模型的新方法。验证结果表明,遗传算法优化后的神经网络模型所需要考虑的参数较少,应用方便,预测精度和效率较高,在城市给水系统水质模拟预测研究中有一定的参考应用价值。Considering with global searching ability of genetic algorithm and partial learning ability of BP neural network, a new method was applied for rapid and efficient establishment a water quality model of residual chlorine in water supply network. The verification showed that neural network had the features of less considerable parameters, convenient for application, high accuracy prediction by a genetic algorithm optimization, and it had certain reference value on simulation and prediction for urban water supply system.

关 键 词：给水管网 遗传算法 神经网络 水质模型 遗传算子 MATLAB工具箱 

分 类 号：TU991.33[建筑科学—市政工程]