摘要
采用动量法和学习自适应调整策略改正的BP神经网络建立了工程造价和主要工程量估算数学模型。针对以往估算模型精度不高的原因,模型中增加了工程造价指数和工程硬、软件环境对工程造价的影响。采用m atlab6.1计算软件,以深圳市已完典型工程资料为例,验证了该模型的正确性及实用性。
A BP network improved by momentum and learning and self-adapting methods is presented to estimate the cost and main quantities,in which the engineering cost index and engineering hardware and software environment are taken into account to improve the precision.Matlab 6.1 is employed to analyze the evaluation model.It is accurate and valuable in practice,as testified by the prototype data from Shenzhen city.
作者
任宏
周其明
Ren Hong;Zhou Qiming(Chong Qing University,Chong Qing 400045,China)
出处
《土木工程学报》
EI
CSCD
北大核心
2005年第8期135-138,共4页
China Civil Engineering Journal
关键词
神经网络
BP神经网络
工程造价
快速估算
估算模型
neural network
BP neural network
engineering cost
quick estimation
estimation model
