论文部分内容阅读
应用神经网络理论，本文提出了圆弧破坏和楔体破坏的边坡安全系数估计的新方法。为解决安全系数估计的知识的学习问题，提出了一种推广学习算法。用它对收集到的边坡实例进行学习，然后进行推广，预测出新边坡的安全系数。与极限平衡法和极大似然法的估计结果进行了比较，可以看出，神经网络方法具有推广预测精度高、自学习功能强、考虑不确定性能力强等特点。
Using neural network theory, this paper presents a new method of slope safety factor estimation for circular arc damage and wedge failure. In order to solve the learning problem of safety factor estimation, a generalized learning algorithm is proposed. Use it to learn the examples of slopes collected, and then promote them to predict the safety factor of new slopes. Compared with the results of the limit equilibrium method and the maximum likelihood method, it can be seen that the neural network method has the characteristics of high accuracy of prediction and promotion, strong self-learning function and strong ability of considering uncertainties.