摘要
在工程估价领域的研究中,利用神经网络进行工程造价快速估算,并取得了一定的研究成果。但是真正应用于实际的不多,取得很好效果的更少,神经网络的设计,对训练方法、因素的选择都会影响实际应用效果。为快速提供企业经验数据,根据神经网络原理和工业厂房造价估算的特点,设计了一个基于BP算法的工业厂房造价快速估算神经网络模型,并对样本选取、特征因素确定及处理等关键问题进行了分析。经仿真训练和验证,表明精度能满足要求,最后把训练得到的知识应用于实际,取得了预期的效果。
In the field of engineering cost estimation,many scholars are engaging in research of cost estimation by using BP neuralnetwork and have gained some study fruits.But there are few applied to practical because of the difficulties in network design,training way and factor selecting.According to the basic principles of neural network and characteristics of industrial plants cost estimation,the model of fast cost estimation of industrial plants based on BP neural network is designed in the paper.Also the paper analysis the problems of swatch selecting,character factor ensuring and disposing.Training and validating results show that the precision meets the estimation requirements.The knowledge gained from training can be applied to practice.
出处
《计算机仿真》
CSCD
北大核心
2010年第8期307-310,367,共5页
Computer Simulation
基金
重点学科建设项目(J51302)
关键词
工业厂房
造价估算
神经网络
Industrial plants
Cost estimation
Neural network