基于灰色RBF神经网络模型在建筑物沉降预测中的应用  

Application of neural network model based on the gray RBF to the building settlement prediction

作　　者：马天驰[1] 

机构地区：[1]黑龙江工程学院测绘工程学院,黑龙江哈尔滨150050

出　　处：《黑龙江工程学院学报》2016年第2期5-7,共3页Journal of Heilongjiang Institute of Technology

摘　　要：建筑物沉降的因素多种多样,因此,采用变形监测技术对沉降量进行监测,利用精度较高的预测模型进行沉降量预测预警很有意义。为了提高建筑物变形监测的精度以及变形监测预警的准确性,针对GM(1,1)模型和RBF神经网络进行分析,提出灰色RBF神经网络模型。通过对某建筑物的沉降监测数据进行实例计算,结果显示灰色RBF神经网络模型具有较高的预测精度。Building settlement’s factors are varied ,so this paper uses deformation monitoring technology for building settlement monitoring and the high precision prediction model for settlement prediction .For the problem of low precision of building deformation monitoring ,this paper analyzes the gray model and RBF neural network and proposes an improved gray RBF neural network model .The calculation results show that the improved gray RBF neural network model has higher accuracy .

关 键 词：灰色模型 RB F神经网络 建筑物沉降 预测 

分 类 号：TU196.1[建筑科学—建筑理论]