基于GM(1,1)+BP神经网络组合模型的路用性能预测研究  

Road Performance Prediction Based on GM(1,1)+BP Neural Network Combined Model

作　　者：苏卫国[1] 吴启槟 王景霄 SU Weiguo;WU Qibin;WANG Jingxiao(School of Civil Engineering and Transportation,South China University of Technology,Guangzhou 510640,China)

机构地区：[1]华南理工大学土木与交通学院,广东广州510640

出　　处：《昆明理工大学学报（自然科学版）》2022年第1期147-155,共9页Journal of Kunming University of Science and Technology(Natural Science)

基　　金：国家自然科学基金项目(51808228)。

摘　　要：为缓解我国公路庞大养护需求与资金有限的矛盾,需从网级角度对路网内所有路段进行决策优化,其中规划期内路用性能预测尤为重要.在分析常用路用性能预测模型特点的基础上,提出GM(1,1)+BP神经网络组合预测模型,即首先利用GM(1,1)对相关路用性能指标进行初步预测;然后根据道路路面属性数据并利用BP神经网络对初步预测结果修正优化,使得路用性能预测更符合路用性能衰减规律;最后通过某市国省道路用性能数据和路面属性数据,验证了GM(1,1)+BP神经网络组合预测模型的可行性与准确性.组合预测模型可作为路用性能预测的有效手段,为科学养护决策提供依据.In order to alleviate the contradiction between China’s huge highway maintenance needs and limited funds, it is necessary to optimize all road sections in the road network from a network-level perspective, and road performance prediction during the planning period is particularly important. Based on the analysis of the characteristics of commonly used road performance prediction models, this research proposes a GM(1,1)+BP neural network combination prediction model. First, we used GM(1,1)to make a preliminary prediction of relevant road performance indicators, and then used BP neural network to modify and optimize the preliminary prediction results according to the road surface attribute data, so that the road performance prediction is more in line with the road performance attenuation law. Last, the feasibility and accuracy of the combined prediction model of GM(1,1)+ BP neural network are verified through the road performance data and road attribute data of a city’s national and provincial highways. It can be used as an effective means of road performance prediction and provide a basis for scientific maintenance decision-making.

关 键 词：路用性能 GM(1 1)灰色模型 BP神经网络 组合预测模型 

分 类 号：U414[交通运输工程—道路与铁道工程]