岩层移动监测是研究开采沉陷问题最有效的方法。由于监测时间长、地表积水、人为损坏等原因,监测点极易缺失。研究缺失监测点的数据插补模型对提高岩移监测数据的利用率、总结开采沉陷规律等都具有重要意义。提出了采用BP神经网络模型进行缺失监测点的数据插补的思路,并采用Matlab实现了该模型。研究结果表明,人工神经网络模型能够很好的逼近地表沉陷盆地,用此模型作为岩移监测点数据插补是完全可行的。 Rock movement monitoring is the most effective method for studying subsidence. Due to the long monitoring time, surface water, man-made damage and other reasons, monitoring points can easily be missing. Studying the data interpolation model of missing monitoring points is of great significance to improve the utilization rate of monitoring data and summarize the law of mining subsidence. The idea of ​​using BP neural network model for data interpolation of missing monitoring points is proposed. The model is implemented by Matlab. The results show that the artificial neural network model can well approximate the surface subsidence basin, and it is feasible to use this model as the data interpolation of the rock migration monitoring points.