摘要
首先构建智慧高速公路利益相关者认知及可持续运营效益的测量指标,然后建立融合One-Hot编码和深度神经网络(DNN)的模型(OH-D模型),以帮助各利益相关者在项目建设前期高效识别智慧高速公路项目可持续运营效益(包括运输效率、服务水平、相关产业发展水平、生产要素流动效率),并在建设、运营过程中采取有效的组织措施和管理对策.结果表明:该模型能根据利益相关者认知识别智慧高速公路项目可持续运营效益,其准确率达96.1%;同时,准确性、收敛速度均优于传统DNN、极端梯度提升(XGBoost)、K-最近邻(KNN)及支持向量机(SVM)等模型.所提模型为立足利益相关者视角识别智慧交通基础设施项目运营效益提供了分析工具和决策支持.
The measurement indicators for stakeholder cognition and sustainable operational benefits of intelligent highway projects are constructed,and then a model that integrates One-Hot encoding and deep neural network(OH-D model),is constructed,to identify the main sustainable operational benefits of intelligent highway projects(including transportation efficiency,service level,development level of related industries,and flow efficiency of production factors)and assist various stakeholders in taking effective organizational and management measures during the construction and operation process.The results show that the model can identify the operational benefits of intelligent highway projects based on stakeholder cognition,with an accuracy of 96.1%.Meanwhile,the accuracy and convergence rate are better than those of the traditional DNN,extreme gradient boosting(XGBoost),K-nearest neighbor(KNN),and support vector machines(SVM).The proposed model provides analytical tools and decision support for analyzing different operational benefits of transportation infrastructure projects from the perspective of stakeholders.
作者
刘瑞
袁竞峰
庞犇
薛斌
Liu Rui;Yuan Jingfeng;Pang Ben;Xue Bing(School of Civil Engineering,Southeast University,Nanjing 211189,China;School of Public Policy and Administration,Chongqing University,Chongqing 400044,China)
出处
《东南大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2023年第6期1119-1127,共9页
Journal of Southeast University:Natural Science Edition
基金
国家自然科学基金重点资助项目(72134002)
国家自然科学基金面上资助项目(72072031)
江苏省研究生科研与实践创新计划资助项目(KYCX22_0319).