摘要
为了研究地铁深基坑墙体偏斜量随时间变化的时空效应规律,对现有的监测数据进行分析,建立GA-SVM预测模型。首先采用三次样条插值法进行数据处理,然后利用GA算法对SVM中的参数进行迭代寻优,将建立好的模型对测试样本进行训练,最后将模型应用在某实例中,研究表明该预测模型的预测结果具有较强的泛化能力和较高的精准度。
In order to study the temporal and spatial effects of the skew amount of the deep foundation pit with time.The existing monitoring data can be analyzed,a GA-SVM model was proposed.First,the measured data ware processed by the cubic spline interpolation method.Then using the GA to optimize the parameters of support vector machine.Training the test samples with the established model,and the model was applied to a practical example.The prediction results have strong generalization ability and high precision.
作者
赵建钗
曹荣
刘俊娥
石祥锋
ZHAO Jian-chai;CAO Rong;LIU Jun-e;SHI Xiang-feng(School of Management Engineering and Business,Hebei University of Engineering,Handan 056038,China;Architectural Engineering College,North China Institute of Science and Technology,Beijing 101601,China)
出处
《价值工程》
2018年第9期147-149,共3页
Value Engineering
基金
2017年安全生产重特大事故防治关键技术科技项目(zhishu-0013-2017AQ)
关键词
深基坑
墙体偏斜量
支持向量机
遗传算法
网格搜寻法
三次样条插值法
deep foundation pit
wall skew amount
support vector machine
genetic algorithm
grid search
cubic spline interpolation method
