基于改进爬山聚类法的模糊神经网络边坡稳定性判别模型

摘要

The factors which control and affect the slope stability are random and fuzzy .Considering the main factors in-fluencing the slope stability ,such as weight ,cohesion ,angle of internal ,angle of slope and the height ,a fuzzy neural network model was established to predict slope stability .The modified mountain clustering method was used for structural study ,and the BP algorithm and least squares estimation algorithm were used to adjust the parameters of the fuzzy neural network model .The results show that the proposed method is feasible and effective in predicting slope stability .%影响边坡稳定性的因素复杂且具有随机性和模糊性。综合考虑重度、黏聚力、内摩擦角、坡角及坡高等影响边坡稳定的主要因素，为判别边坡稳定性建立出新型模糊神经网络模型，该模型利用学习能力强大的神经网络及推理功能突出的模糊逻辑，通过改进的爬山聚类法进行结构学习，并利用 BP 算法和最小二乘估计法相结合的综合学习算法来调整参数，进而大幅度提高模型判别能力。经工程实例测试证明该模型可以快速准确的判别边坡的稳定性，可以为类似工程提供参考和借鉴。