中文摘要：

本文通过对BP网络模型的研究，建立BP神经网络预测模型，并使用VC＋＋语言编程加以实现。将模型应用于东江大坝实测变形监测数据的分析和预测，对时间序列建模和以环境变量为自变量两种方法建模，发现这两种模型都可以很好的实现变形量的预测，且以自变量为因素所建立的模型预测精度更高。

英文摘要：

This paper focuses on the research of the BP neural network models. The BP forecasting models have been established, and carried out with the VC ＋ ＋ program for the purpose of analyzing and forecasting the real time monitoring data of the Dong Jiang dam real time. Generally, there are two methods to establish the model. One is based on a continual time series, the other is to use en- vironment variables as independent variables. In the paper, both methods can get good forecasting results, and the accuracy of the sec- ond method is higher than that of the first one.