摘要
介绍了现有的BP神经网络在预测方面的应用实例,建立了BP网络的预测模型,结合都江堰渠首岷江上游来水量的预测,进行了仿真实验。实验结果表明,该模型是可行的,并且具有较高的预测精度。在模型可行的基础上,分析了LMBP算法、自适应学习速率动量梯度下降反向传播算法、弹性反向传播算法三种BP神经网络训练算法得到的仿真结果的差异,进而讨论了BP神经网络采用不同的BP网络训练算法对预测结果的影响。
In this paper the performance of back propagation neural networks applied to the predicting problems is evaluated. Apply-ing the designed BP networks model to the Dujiangyan water time series predicting problems,a great deal of experiments and simulations show that the model and its algorithm is feasible and the different results by three different methods (LMBP Algorithm,momentum gradient descent BP algorithm,Flexible BP algorithm).
出处
《微计算机信息》
2009年第34期149-150,共2页
Control & Automation