【摘要】 ＢＰ网络是目前应用很广的人工神经网络。改善ＢＰ网络算法尤其是提高ＢＰ网络的学习速度一直是人们研究的重点课题。本文在对ＢＰ网络现行算法研究的基础上，提出根据系统误差自适应调节学习率的快速ＢＰ算法（ＦＢＰ），并以棉花棉铃虫预测为例，对两种算法的收敛速度进行了比较。更多还原

【Abstract】 BP network is a widly applicated Artificial Neural Network. To improve BP algorithm, especialy to raise its learning speed, is all long a very important reserched ob-ject. This paper advances a Fast BP algorithm (FBP). It accords to the system error to adjust learning speed. By procasting cotton heliothis armigera, we made a compare to these two kinds BP algorithm’s convergence speed.更多还原