【摘要】 在分析影响滑带土强度因素的基础上,建立了滑带土强度参数的 BP 神经网络模型,预测滑带土在不同含水率下 c,φ值的变化规律,尤其是当红石包滑坡前缘、后缘地质条件差异较大时,找出可能的工况匹配,可以为滑坡稳定性评价提供可靠依据,克服了 c,φ值按峰值折减的主观性。应用表明:该模型精度很高,有应用前景。更多还原

【Abstract】 Based on analyzing the factors effected on the strengthes of slide zone soil,a BP neural network model is established. With this model a series of changing rules of the value of cohesion(c) and the angle of internal friction(Φ) are predicted in different water content;especially in Hongshibao landslide where the geological conditions are very different between front and rear.The match of the values ofc and Φ gained by this model builds a firm foundation for the further stability evaluation;and the subjectivity existed in the method of discount of the values of c and Φ is effectively avoided.Its practice shows that the model has a higher precision and a good perspective.更多还原