基于Logistic回归的水利工程边坡稳定性评价

Evaluation of slope stability of water conservancy project based on logistic regression

水利工程建设中因场地的限制,往往需要对山体进行开挖和填补,因此自然边坡和人工开挖边坡的稳定性是影响工程质量的关键因素。在可行性研究阶段,进行简单可行的坡体稳定性评价十分必要。基于Logitsic回归的思想,建立水利工程边坡稳定性评价模型。为充分提升评价模型的准确性,以坡高、坡角、地层岩性、岩体结构、软弱地层、风化程度、地下水和年降水量等8个主要因素为自变量,坡体稳定性为因变量,以最速梯度下降法作为参数迭代方法,经训练集对模型的训练,得到了水利工程边坡稳定性评价模型,运用于实际工程检验了模型的准确性与实用性,对水利工程边坡稳定性的初步评价工作具有较大的借鉴意义。

It is often necessary to excavate and fill the mountain due to the limitation of the site in the construction of water conservancy projects, so the stability of natural slope and artificial slope excavation is the key factor affecting the quality of the project.It is necessary to evaluate the stability of slope in the stage of feasibility study. A slope stability evaluation model for hydraulic engineering is established based on Logistic regression. Eight major factors of high slope, slope angle, formation lithology, structure, weak formation, weathering degree of rock mass, groundwater, annual precipitation and so on, are considered as independent variables in order to fully improve the accuracy of the evaluation model, and the slope stability is regarded as the dependent variable. The steepest gradient descent method is adopted as the parameter iteration method. The training set is used for training the model, the water conservancy project slope stability evaluation model is obtained, the engineering example is applied to test the accuracy and practicality of the model, and it has great significance to the preliminary evaluation of the water conservancy project slope stability.