水利枢纽下游河段设计最低通航水位推算方法探讨

在分析影响水利枢纽下游河段设计最低通航水位因素的基础上，将设计最低通航水位计算问题概括为一种非线性输入输出的泛函关系，以人工神经网络原理建立推算水利枢纽下游河段设计最低通航水位的BP神经网络模型，用此模型，对闽江水口坝下观音岐设计最低通航水位进行网络计算，在此基础上，将BP网络模型与模拟分析进行了对比分析，2种方法所得结果相近， 但BP网络模型具有较高精度，本文为水利枢纽下游设计最低通航水位计算提供了一种新方法。

Study on method of calculating designed lowest navigable stage at downstream stretch of hydro-junction

Based on analyzing affecting factors of the designed lowest navigable stage, the problem of calculating the designed lowest navigable stage at downstream stretch of hydro junction is generalized as a nonlinear functional relation between input and output. The back propagation (BP) networks mode calculating the designed lowest navigable stage is established on artificial neural network model principle. The designed lowest navigable stage at Guanyingqi downstream Shuikou dike of Minjiang is calculated by the BP network mode. A comparison between BP network model and the model analysis on observed data is made. The results of two methods are about the same, but the BP network model possesses higher accuracy. Thus a new method is provided for the design of lowest navigable stage at downstream stretch of hydro junction.