LS-SVM在桥梁结构健康预测评估中的研究

Study on LS-SVM in Evaluation of Bridge Structural Health Monitoring

桥梁的结构变形包含的桥梁结构内涵信息丰富,具有非线性、时序性和样本容量小的特点.利用支持向量机,以杭州湾大桥实测变形数据为研究对象,提出了基于LS-SVM的桥梁结构变形预测模型,通过实验证明了运用其对桥梁结构健康状况进行评估的可行性和有效性,并且通过实验结果对比显示了最小二乘支持向量机在变形预测中的优势.

Structural distortion of bridge contains rich connotation information of bridge structures and possesses the features of nonlinear,sequential and small sample capacity.In this article,the calculated model of bridge structural distortion is given out by application of support vector machine and the practical monitoring data of Hangzhou Bay Bridge are taken as the study object.The feasibility and effectiveness of evaluation over bridge structural health state have been proved through test,and the superiority of least square support vector machine in distortion calculation has been shown on comparison of test results.