基于改进C-C方法相空间重构和LS-SVM的隧道拱顶沉降预测模型

The Prediction Model of Metro Vault Settlement Based on Developed C-C Method Phase Space Reconstruction and LS-SVM

在地铁工程的设计、施工、工后沉降控制过程中,拱顶下沉监测值是反映地下工程结构安全和稳定的重要数据.针对常用的地铁拱顶沉降测模型只能做短期预测,精度不高,且需要一些土的本构参数的问题,将相空间重构、最小二乘支持向量机理论相耦合,建立基于改进C-C方法相空间重构和最小二乘支持向量机的地铁隧洞拱顶沉降混沌时间序列预测模型.经实例演算,模型比传统C-C方法相空间重构、基于最大Lyapunov指数的混沌预测模型、人工神经网络模型拟合效果好,预测精度高.

The values of metro vault settlement are considered as important data reflecting the safety and stability of the underground engineering structures in the process of design,construction,settlement control of the subway construction.Considering that the common prediction models of metro vault settlement can only make short-term forecasting and need a large number of sample data,a new model are created,which combined phase space reconstruction,the LS-SVM and the chaos theory.The examples proves that the new model is superior to the chaotic prediction model based on the traditional C-C method,the maximum Lyapunov-index and the rbf neural network prediction model.It enriches and develops the metro vault settlement prediction model theories.