摘要
针对城市河网缺乏足够的实测资料和河网水动力学模型模拟速度慢的特点,提出将河网水动力学模型与遗传算法、神经网络方法结合,建立河网智能模型。模型中,利用河网水动力学模型提供神经网络所需的信息,遗传算法用于优化神经网络的初始权重。将该模型应用于上海市浦东新区河网中,智能模拟结果与经过实测资料验证的河网水动力学模型的模拟结果吻合较好,表明河网智能模型精度与水动力学模型接近。同时实时性较好,可用来预测河网水位变化特性,也为今后类似研究提供一种模拟技术。
Considering the lack of the sufficient measured data of the cities' river net and unsatisfactory simulation speed of the hydrodynamic model, an intelligence model for river net is presented in this paper combining the hydrodynamic model with genetic algorithm and artificial neural networks. The information is supplied by the hydrodynamic model and the original weights of artificial neural networks are optimized by the genetic algorithm. The model is used to simulate the river net of Pudong New District in Shanghai. Sound agreement is obtained between the results of intelligence model and those of hydrodynamic model. It shows that the accuracy of intelligence model for river net is close to that of the numerical model, and the intelligence model has an advantage of good real time performance. It provides a good technique for forecasting water level and the similar problems.
出处
《水科学进展》
EI
CAS
CSCD
北大核心
2008年第2期232-237,共6页
Advances in Water Science
基金
新世纪优秀人才支持计划资助项目(NCET-04-0494)
国家自然科学基金资助项目(50479068)~~
关键词
河网
智能模型
人工神经网络
遗传算法
水动力学模型
river net
intelligent model
artificial neural networks
genetic algorithm
hydrodynamic model