基于模糊神经网络的建筑工程造价预测模型  

Predicting Model for Construction Engineering Cost Based on Fuzzy Neural Network

作　　者：梁喜[1] 刘雨[1] 

机构地区：[1]重庆交通大学经济与管理学院,重庆400074

出　　处：《技术经济》2017年第3期109-113,共5页Journal of Technology Economics

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.

关 键 词：建筑工程 工程造价 造价预测 模糊神经网络 

分 类 号：TU723.3[建筑科学—建筑技术科学]