改进BP神经网络在木构古建筑中的寿命预测

标准的BP神经网络存在训练速度慢、容易陷入极小点、泛化能力低的特点。文中用附加动量项和改进学习速率相结合的方法对标准的BP神经网络进行了改进，并将其应用在木构古建筑的寿命预测中。仿真结果表明，和标准的BP神经网络相比，改进后的BP神经网络提高了泛化能力，能较准确地拟合训练值，避免了在确定计算参数过程中所产生的计算误差。

Prediction of Wooden Ancient Building Life by Improved BP Neural Network

Standard BP neural network has characteristics of slow training speed,falling into the minimal point easily and low generaliza-tion ability. In this paper,the standard BP neural network has been improved with additional momentum term and improved learning rate, and applied into the forecasts for the life-span of wooden ancient buildings. Simulation results show that the generalization ability of the improved BP neural network has improved compared with the standard BP neural network,the improved BP neural network can be more accurately fit for the training value,and avoid the calculation error in the process of determining the calculation parameters.