摘要
利用直觉模糊分析方法求得工程特征对建筑工程造价的影响程度系数,据此确定了5个对建筑工程造价影响较大的工程特征,并将之作为神经网络模型的输入向量,构建了基于模糊逻辑和神经网络方法的建筑工程造价预测模型。结合MATLAB工具箱对程序进行设计,并用已完工程数据对模型进行训练、修正以及实例验证,证明了该模型可以有效提高建筑工程造价预测的精确度,具有较强的实用价值。
This paper calculates the influence degree coefficients of engineering characteristics which reflect the relationship between engineering characteristics and construction engineering cost.According to the result,it identifies five engineering characteristics which have greater influences on construction engineering cost,as the input vector of neural network model.Combining with MATLAB toolbox to design program,and using the engineering data to train,fix and verify the model,it proves that this model could effectively improve the accuracy of prediction for engineering cost,and has strong practical value.
出处
《技术经济》
CSSCI
北大核心
2017年第3期109-113,共5页
Journal of Technology Economics
关键词
建筑工程
工程造价
造价预测
模糊神经网络
construction engineering
engineering cost
cost estimating
fuzzy neural network
