道路工程SOM-RBF神经网络估价模型研究

研究从道路工程全生命周期造价的角度,运用自适应(SOM)和径向基(RBF)神经网络的模型与全生命显著性造价(WLCS)结合的预测模型对已完工程造价信息进行数据挖掘,估算工程造价。建立WLCS模型并分析计算。运用自组织SOM网络对WLCS模型分析结果进行特征抽取,验证其对类似工程的确定。在此基础上,运用基于自组织RBF网络对投资估算进行预测。结果表明,WLCS应用可以减少造价估算的工作量,SOM网络可以对道路工程的显著性成本项目(CSIs)造价数据进行特征提取,WLCS的SOM-RBF网络模型减少了估价计算量,并提高了估价精度。本研究成果适用于道路工程造价管理研究和应用领域。

Research on Estimating Models of Whole Life Cost for Road Engineering Based on SOM-RBF Neutral Network

This paper stood by the view of Whole Life Cycle management.The mining of completed project cost data was based on Self-Organization Mapping(SOM)and Radial Basis Function(RBF)models combined with Whole Life Cost-significant(WLCS)for project cost estimation.The WLCS model was established and analyzed.The SOM network was used to extract features from the analytical results of the WLCS model to verify its determination of similar projects.On this basis,the RBF network was used to forecast cost investment.The results show that the application of WLCS can reduce the workload of cost estimation.The SOM network can extract the characteristics of cost data of Cost-significant Items(CSIS)of road engineering.The SOM-RBF network model reduces the amount of calculation and improves the accuracy of estimation.The research results are applicable to the research and application of road engineering cost management.