摘要
丹江口大坝加高后,回水顶托对黄龙滩电厂产生了一系列不利影响,在分析这些不利影响的过程中,需要计算丹江口大坝到黄龙滩电厂的水位流量关系。而传统的水力学方法计算量大,且过程复杂。为了能够简单高效地进行水位流量关系的计算,提出了一种基于BP神经网络模型的方法,将丹江口水库到黄龙滩电厂的河道分为两段,构建两个BP神经网络模型分别进行模拟计算。结果表明:丹江口水库至堵河口的模型的水位模拟结果最大偏差为0.011 m,均方根误差为0.0033 m;堵河口至黄龙滩电厂的模型的水位模拟结果最大偏差为0.2469 m,均方根误差为0.0841 m。两个模型的模拟精度都较高,证明该方法具有可行性。
Since the heightening of Danjiangkou Dam,backwater of the reservoir has caused a series of negative effects on Huanglongtan Hydropower Plant.To analyze those effects,it is necessary to get the stage-discharge relationship of the river channel between Danjiangkou Dam and Huanglongtan Hydropower Plant.Considering that the traditional hydraulic method is usually of large amount and complicated computation work,a simple and efficient method based on BP neural network model is proposed.The channel between Danjiangkou Dam and Huanglongtan Hydropower Plant is divided into two sections,and two BP neural network models are built and simulated.The results show that the maximum deviation and root mean square error of the first model of the river channel from Danjiangkou Dam to Duhekou are 0.011 m and 0.0033 m respectively.For the second model of the river channel from Duhekou to Huanglongtan Hydropower Plant,they are 0.2469 m and 0.0841 m respectively.It shows that the proposed method is feasible and effective as the results are satisfactorily accurate.
作者
袁林山
张力
YUAN Linshan;ZHANG Li(State Grid Hubei Huanglongtan Hydropower Plant, Shiyan 442000, China)
出处
《水电与新能源》
2020年第1期7-11,共5页
Hydropower and New Energy
关键词
BP神经网络模型
水位流量关系
黄龙滩电厂
BP neural network model
stage-discharge relationship
Huanglongtan Hydropower Plant