In this paper,the 18 S province highway engineering cost data information are collected,and as a basic data information for building highway engineering cost based on fuzzy neural network prediction model for the corresponding test and training.The highway engineering cost prediction model to quantify the engineering data collected,the form of the all materials are numerical,then the method of normalization,unified handling these values,the final reference to relevant basic knowledge of fuzzy mathematics theory,the screening of the corresponding sample project,and calculate the degree of press close to,the network training samples for scientific choice,then plug in processed data in the model,the output is obtained.This article g example fully proved that the proposed highway project cost prediction model has stronger generalization ability,and has obtained the remarkable effect in the real engineering,even not enough sample database can obtain data of highway engineering cost reasonably.%对S省境内18条高速公路工程造价数据资料信息进行收集,并作为基础数据信息用于构建基于模糊神经网络的高速公路工程造价预测模型进行相应的测验及训练.该高速公路工程造价预测模型先要量化所搜集的工程资料,所有资料的存在形式均为数值,随后再采用归一化方法,统一处理这些数值,最终借鉴模糊数学相关基础知识理论,筛选相应的样本工程,并计算贴近度,将网络训练样本进行科学选择,随后在模型当中代入经过处理的数据,进而得到输出结果.实例验证充分证明,提出的公路工程造价预测模型具有较强的泛化能力,并且在现实工程中取得了显著的效果,即使样本数据库不够丰富也能获得合理的公路工程造价数据.
展开▼
