关键词: BP神经网络, 多元回归, 边坡稳定, 预测

Abstract: Based on the main natural and engineering factors affecting on slope stability, BP artificial neural network was applied to predict the stability of rock slope. After network training by a great deal of swatches, the reliable BP network model was elicited to forecast the optimum angle of final slope of mine, and the results were compared with the ones draw from multiple regression analysis. The study showed that the method of artificial neural network has characteristics of high precision, convergence rapidity and high fault tolerance capability, which can meet the need of slope stability prediction in engineering practice.

Key words: BP artificial neural network, Multiple Regression, Slope stability, Forecasting