Abstract: Dumping site landslide is one of the major disasters for mine,threaten the safety mining of mine.The early warning index system for mine dumping site landslide is proposed,according to the comprehensive early warning indexes,the dumpling site landslide early warning model based on BP neural network is established.The indexes including adhesion,internal friction angle,surface displacement,internal displacement,rainfall etc.are selected as the input units of BP network,appropriate excitation function and learning step are selected,and 20 sets of sample data for actual engineering are also selected for conducting the training and learning of BP neural network by using improved gradient algorithm.The learned and trained early warning model is applied to analyze the landslide of the dumpling site of a mine in Guizhou Province,the early warning analysis results is accordant with the actual situations of the dumping site,which show that the early warning model established in this paper is reasonable and has application value.

Key words: Dump site landslide, Early warning index, Early warning model, BP neural network