摘要
传统建筑工程成本分析方法存在运行效率低、收敛性能差,导致成本分析不准确。因此提出基于粗糙集理论的建筑工程成本分析方法,构建粗糙集-小波神经网络预测模型,实现建筑工程成本的准确分析,通过粗糙集属性约简过滤掉冗余属性,减少小波神经网络输入节点,降低网络结构的复杂性,提高网络训练效率和精度。利用此预测模型,将采集到的建筑工程成本干扰因素通过粗糙集理论实施约简处理,将约简的因素作为小波神经网络的节点输入训练网络,得到建筑工程成本的分析结果。实验结果说明,所提方法具有较高的运行效率和收敛性能,能够对建筑工程成本进行快速、准确的分析。
The traditional cost analysis method of construction engineering has low operation efficiency and poor convergence performance,which leads to inaccurate cost analysis.Therefore,a construction project cost analysis method based on rough set theory is put forward,in which the rough set and wavelet neural network forecasting model is constructed to realize the accurate analysis of the cost of construction engineering.The attribute reduction of rough set can filter out the redundant attributes to decrease the number of input node of wavelet neural network,reduce the complexity of network structure,and improve the efficiency and precision of network training.The acquired factors influencing the construction engineering cost are reduced by means of rough set theory,and the reduced factor is deemed as input node of wavelet neural network for training network to get the analysis result of construction engineering cost.The experimental results show that the proposed method has high operation efficiency and convergence performance,and can quickly and accurately analyze the cost of construction project.
作者
侯文婷
HOU Wenting(Wuhan University of Science and Technology,Wuhan 430081,China;Inner Mongolia Technical College of Construction,Hohhot 010050,China)
出处
《现代电子技术》
北大核心
2018年第19期83-88,共6页
Modern Electronics Technique
关键词
粗糙集理论
建筑工程
成本分析
小波神经网络
约简
冗余属性
rough set theory
construction engineering
cost analysis
wavelet neural network
reduction
redundant attribute
