基于改进BP神经网络理论的算法研究

研究了改进标准BP神经网络的方法，特别是研究加快标准BP神经网络收敛速度的方法。针对标准BP神经网络的收敛速度慢，分别采用3种方法修改网络的学习速率，然后建立了相应的改进BP网络模型，并把模型应用到实际生活中。和标准BP网络相比，提出的3种改进模犁都具有收敛速度快、精度高的优点。

Research of Algorithms Based on Improved BP NN Theory

Methods of improving the standard BP network are researched in the dissertation, and methods of quickening the convergence speed of standard BP network are researched Aiming at especially the slow convergence speed of standard BP network, three methods are used respectively to modify the learning rate of network. Models improved by the methods are established, and then applied in practice. Compared with standard BP network, all of improved models have advantages faster convergence speed and higher efficiency.