论文部分内容阅读
该文提出了一种判别水库水温分层模式的人工神经网络方法，该方法不直接描述水温分层的物理机制和物理模型，通过设定前馈网络结构并利用已有实例进行训练，使网络能很好地模拟水温垂直分层模式影响因素与水温分层模式间的映射关系，其特点是分类准确、适应能力强，对线性和非线性分类问题均适用。
In this paper, an artificial neural network method for discriminating reservoir water temperature stratification is proposed. The method does not directly describe the physical and physical models of water temperature stratification. By setting up the feedforward network structure and using existing examples to train, the network Which can well simulate the mapping relationship between the influencing factors of water temperature vertical stratification model and water temperature stratification model. It is characterized by accurate classification and strong adaptability, which is applicable to both linear and nonlinear classification problems.