基于最小二乘法的逻辑回归模型在土石坝沉降预测中的应用

摘要

In accordance with the shortage of the conventional regression model, it is proposed herein that establishing the settlement prediction model of earth-rock fill dam with logistic regression, for which the parameters of the model are determined with the method of least square. The model is applied to the settlement prediction of the dam for Nanwan Reservoir. The study shows that the fitting error of this method is smaller than that from the conventional one, and then better prediction effect can be obtained as well.%针对传统回归模型的不足,提出了采用逻辑回归曲线建立土石坝沉降预测模型,由最小二乘法确定模型参数.该模型用于了南湾水库大坝的沉降量预测.研究表明,该方法的拟合误差较传统模型小,可取得较好的预测效果.