基于BP神经网络的矿山地下水位预测研究

摘要

采用BP神经网络技术,将矿区的降雨量、排水量及前期水位三个因素作为输入层,矿山地下水位作为输出层,建立矿山地下水位预测模型.文章详细介绍了BP神经网络实现矿山地下水位预测的基本算法,将研究区矿山的长期观测孔实测水位作为实验数据并作出误差分析.最终成果能够达到矿山地下水位预测目的,并为分析地下水降落漏斗趋势提供有力依据.%The model of the prediction of mine groundwater level is set up adopt the BP neural network technology, use the three factors of the rainfall, displacement and pre—water level of the mining area as the input layer, as well as use mining groundwater level as the output layer. This paper describes the basic algorithm of how to use BP neural network realize prediction of mining groundwater level, the measured water level of the long—term observation wells in the mine of the study area as the experimental data and to make the error analysis. The final outcome can be achieved the purpose of predict mining groundwater level, and provides a strong basis to analyze the trend of groundwater depression cone.