改进的标准模糊神经网络的工程造价快速估算

Fuzzy Neural Network Based on Improved Standard Model and Its Application in Calculating Project’s Cost Quickly

在对影响建筑工程造价因素分析和标准模糊神经网络结构分析的基础上,通过增加输入层与模糊层之间的权值v,加入规则的重要度γ,对标准模糊神经网络进行了改进,并建立基于改进的标准模糊神经网络的工程造价快速估算模型。将基于这种结构的模糊神经网络的工程造价快速估算模型应用于建筑工程的投标报价中,从仿真结果可以看出该网络模型学习时间较短,学习速率较快,精度较高。

Considering the reasons which can influence the project＇s cost and the fuzzy neural network architecture, the weight v between the input layer and the fuzzy layer, and the importance of rules were added to improve the fuzzy neural network based on the improved standard model and an model of calculating the project＇s cost quickly was constructed based on the improved network. At last, the data of the similar projects was used to train the fuzzy neural network by simulation, and some more exactly forecastable results were obtained. Synchronously the idea of calculating the project＇s cost quickly was made come true. The results of simulation indicate the method is feasible.