论文部分内容阅读
采用电化学腐蚀的方法加速梁内钢筋锈蚀,使用分级加载方案对受腐蚀钢筋混凝土梁进行了抗弯承载力试验研究,获得了20根不同锈蚀率的钢筋混凝土梁承载力极限值.使用现有的试验数据构成训练样本对,供支持向量机进行预学习.在机理建模方法中,基于线性核函数分别获得了分段式和整体式的回归公式;在黑箱建模方法中,基于径向基核函数获得了黑箱预测模型.通过将预测结果和本文试验值进行对比,验证所提出的基于支持向量机的受腐蚀混凝土梁承载能力预测方法有效性;并将预测结果与其它回归公式对比,表明所提出的预测方法具有预测精度高、适用范围广和预测策略多等优点.
Electrochemical corrosion method was used to accelerate the corrosion of the steel bars in the beam. The load-bearing capacity of the reinforced concrete beams subjected to corrosion was investigated by using the graded loading scheme, and the ultimate bearing capacity of 20 reinforced concrete beams with different corrosion rates was obtained. Of the experimental data to form a training sample pair for SVM pre-learning.In the mechanism modeling method, based on the linear kernel function, respectively, obtained a segmented and integral regression formula; in the black box modeling method, based on the radial The kernel function obtains the black box prediction model.Through comparing the prediction results with the experimental data, the validity of the proposed method for predicting the bearing capacity of corroded concrete beams based on support vector machines is verified. Compared with other regression formulas, It shows that the proposed prediction method has the advantages of high prediction accuracy, wide range of application and many prediction strategies.