【摘要】 基坑开挖引起其邻近地表沉降的即时预测评估有利于实现工程安全的高效控制,应用支持向量机(Support Vector Machines,SVM)函数方法提出一个适用于这项工作需要的沉降数据预测方法。通过遗传算法(Genetic Algorithm,GA)的引入获得核函数类型、核函数参数及错误惩罚因子的选取结果。结合SVM函数回归计算技术,利用已知数据完成GA-SVM建模,根据模型的外推结果,给出沉降预测值。以广州某地铁工程为实例,对比分析了GA参数寻优是否进行优化的预测效果差异,与实测结果的对比证实GA-SVM预测模型具有较好的预测精度,同时表明GA-SVM沉降预测方法良好的技术应用价值。更多还原

【Abstract】 Immediate prediction and evaluation of ground settlement caused by foundation pit excavation is helpful to control of engineering safety. In this paper,a settlement data prediction method based on Support Vector machine( SVM) function is proposed for this job need. Genetic algorithm is used to optimize the model parameters,such as kernel function type,the kernel function parameter and error warning factor. GA- SVM modeling can be completed by the known data and the function of SVM regression computing technology. According to the results of the model ’s extrapolation subsidence prediction is given. Comparing with SVM model without parameters optimization and the actual monitoring results by Guangzhou subway project,analysis represents that the genetic optimization based support vector machine prediction model has good accuracy and this model has certain practical value for the safety monitoring of deep foundation pit.更多还原