基于人工神经网络的龙羊峡水库入库径流预报

Forecast to Longyangxia reservoir inflow runoff based on an artificial neural network

阐述神经网络是一个大规模的非线性功能系统,具有很强的非线性处理能力。人工神经网络模型是高度的非线性模型,能够有效地模拟本质为非线性的实际水文系统。又介绍了人工神经网络的结构、原理,并结合龙羊峡水库入库径流资料,利用BP神经网络方法进行预报分析,预测值和实测值的相对误差均在10%以内,预测结果相对较好。

The neural network is a large-scale nonlinear function system with a strong nonlinear processing ability and the artificial neural network model is a high nonlinear model and it can simulate a real nonlinear hydrologic system.The structure and principle of the artificial neural network are introduced in this paper and the forecast analysis to the Longyangxia reservoir inflow runoff data is down using BP neural network method.The forecast result is better relatively and the relative error between forecast value and real value is within 10%.