床面形态判别的人工神经网络方法  

Classification of Bed Forms Based on Artificial Neural Network

作　　者：黄才安[1] 王进 

机构地区：[1]扬州大学水建学院,扬州225009 [2]江都市昭关自流灌区管理处,江都225200

出　　处：《应用基础与工程科学学报》2002年第1期51-56,共6页Journal of Basic Science and Engineering

基　　金：江苏省教育厅自然科学基金资助项目 (0 0KJB5 70 0 2 )

摘　　要：根据给定的水流及泥沙条件 ,预报床面可能出现的沙波形态 ,是一个重要的研究课题 .在对影响床面形态的参数分析后 ,认为床面形态应由 3个无量纲参数控制 ,因而以往采用 2个参数的经验方法有很大的局限性 .基于此结论 ,建立了预测床面形态的人工神经网络模型 ,该模型充分考虑了本问题的特点 .验证结果表明 ,该模型具有较高的精度 。According to given flow and sediment conditions, predicting possible bed form in river bed is an important problem. After analyzing the parameters that affect the bed form, it is believed that bed form is controlled by three dimensionless parameters. So, early empirical methods that used only two parameters to predict bed form have limits. Based on this result, an artificial neural network model which take into account the characteristics of bed form problem has been established to predict bed form. Test results show that this model has a high accuracy, and can be used in practical problems.

关 键 词：床面形态 人工神经网络 BP模型 河床 沙波形态 水流 泥沙 

分 类 号：TV14[水利工程—水力学及河流动力学]