论文部分内容阅读
在边坡稳定性分析中,边坡安全系数受地质条件、地貌因数、水文气候、地震作用、风化作用等众多因素的影响,这给边坡评价带来了极大的困难。采用A-K-GN法预测边坡安全系数:用层次分析法对影响边坡稳定性的主要因素进行分析;用自组织竞争kohonen神经网络对边坡样本进行归类;运用经过遗传算法优化的BP神经网络(遗传神经网络)方法,建立边坡安全系数隐函数关系式,从而预测边坡安全系数。用kohonen神经网络归类后的边坡数据为样本,用层次分析法选取了容重γ、粘聚力c、内摩擦角φ、边坡角α、边坡高度H和孔隙压力比γu作为边坡安全系数隐函数的随机变量输入单元,以边坡安全系数F作为输出单元。通过预测值与实际值的对比分析,验证了A-K-GN法预测边坡安全系数的合理性。
In slope stability analysis, the safety factor of slope is affected by many factors, such as geological conditions, geomorphic factors, hydrological climate, earthquake action and weathering, which bring great difficulties to slope evaluation. The AK-GN method is used to predict the slope safety factor. The analytic hierarchy process (AHP) is used to analyze the main factors affecting the slope stability. The kohonen neural network is used to classify the slope samples. The BP neural network Network (genetic neural network) method to establish the slope safety factor implicit function relationship, and thus predict the slope safety factor. Using slope data classified by kohonen neural network as sample, the gravimetric method was used to select the gravimetric γ, cohesion c, internal friction angle φ, slope angle α, slope height H and pore pressure ratio γu as the slope The safety factor implicit function random variable input unit takes the slope safety factor F as the output unit. By comparing the predicted value with the actual value, the rationality of A-K-GN method in predicting slope safety factor is verified.