大数据驱动的服务运营系统性优化与管理——以新能源汽车为例

项目来源

国家自然科学基金(NSFC)

项目主持人

冉伦

项目受资助机构

北京理工大学

立项年度

2017

立项时间

未公开

项目编号

91746210

研究期限

未知 / 未知

项目级别

国家级

受资助金额

240.00万元

学科

管理科学-工商管理-企业运营管理

学科代码

G-G02-G0211

基金类别

重大研究计划-重点支持项目-大数据驱动的管理与决策研究

关键词

数据驱动 ; 深度商务分析 ; 新能源汽车 ; 服务运营 ; 大数据 ; big data ; data-driven ; service operations ; electric vehicle ; business analytics

参与者

申作军;刘鹏;江连山;高广宇;张玉利;褚宏睿;贾琳

参与机构

清华大学;中创三优(北京)科技有限公司

项目标书摘要:目前大数据技术在数据分析和信息处理方面日趋完善,然而针对服务运营管理与决策的大数据优化理论、方法与实践研究仍亟待深入。本项目旨在建立大数据驱动的服务运营管理与决策理论与方法,并针对制约新能源汽车服务运营效率的关键问题进行应用研究。首先,提出以数据融合、知识融合和决策融合为特点的大数据驱动的服务运营管理与决策范式,构建了基于随机鲁棒优化、机器学习和混合模型的大数据驱动的系统性优化方法。其次,构建涵盖“人—车—桩—网”的新能源汽车大数据集成平台,并研究大数据驱动的用户行为模型识别与需求预测方法。最后,采用提出的大数据驱动的优化方法,对新能源汽车的充换电设施选址与实时运营调度问题进行系统性研究。项目将机器学习、运筹学和运营管理理论及方法结合,为面向大数据的服务运营管理与决策提供理论基础,同时研究成果将为新能源汽车行业规划与管理、企业的战略管理和实时调度提供高效、可靠的科学决策支持。

Application Abstract: Big data technologies have achieved a great progress in data analytics and information processing in the past decade.However,there is a lack of research on big data-driven optimization theory,method,and practice to support management and decision-making in service operations.This project aims to address this gap by developing a big data-driven management and decision-making theory for service operations.The project will also perform a practical research to eliminate bottlenecks that impede the efficient operation of the electric vehicle industry.Specifically,this project first develops a big data-driven management and decision paradigm of service operations that is characterized by data fusion,knowledge fusion and decision fusion,and proposes a systematic big data-driven optimization approach based on stochastic robust optimization,machine learning,and mixture models.Secondly,the project will build a big data platform for the electric vehicle industry that integrates data of drivers,vehicles,EV chargers,and power grid.Using this big data platform,the project will propose big data-driven approaches to identify drivers’behavioral patterns and thereafter forecast their driving demands.Finally,this project will perform a systematic research on how to optimize the electric charging facilities location decisions and the real time operations dispatching using the proposed big data-driven optimization approaches.This project combines machine learning,operations research,and theories of operations management and provides a theoretical basis for big data-driven management and decision making in service operations.Our research outcomes will provide efficient and reliable decision supports to industry-level planning and administration,and firm-level strategic management and real time dispatching.

项目受资助省

北京市

项目结题报告(全文)

本项目聚焦于大数据驱动的服务运营管理与决策问题,结合大数据时代服务运营管理的特点和发展趋势,分别从大数据驱动的服务运营管理与决策的优化理论与方法,大数据驱动的新能源汽车服务运营实践两方面展开研究。研究成果可为面向大数据的服务运营管理与决策提供理论基础,为新能源汽车行业的规划与管理提供高效可靠的科学决策支持。本研究取得了如下成果:(1)提出了大数据驱动的服务运营管理与决策范式,构建了大数据驱动的系统性优化框架;(2)构建了涵盖“人—车—桩—网”的新能源汽车大数据集成示范平台;(3)分析了新能源汽车用户乘车的出行模式和充换电模式以及对应的需求预测;(4)解决了大数据驱动的新能源汽车设施选址及充换电运营调度问题。课题组在国内外主流学术期刊及会议发表论文37篇,其中SCI/SSCI检索期刊论文32篇,包括PNAS1 篇,MS1篇、MSOM3篇、POM3篇、JOC3篇、TS1篇、TRB1篇等。项目共培养毕业博士研究生3名、硕士研究生12名。项目组成员获得多项学术奖励,包括2020年北京市科学技术奖一等奖、2019 informs Conference of Service Science Best Student Paper-The First Place Winner、2019中深杯全国新能源汽车大数据创新创业大赛创新组电动汽车行驶SOC预测赛题金奖、M&SOM Meritorious Service Award等。

  • 排序方式:
  • 1
  • /
  • 1.Optimal Decision and Joint Carbon Reduction in a Supply Chain with Carbon Emissions Dependent Stochastic Demand

    • 关键词:
    • Carbon;Carbon Economy;Carbon emissions;Contracts;Cost effectiveness;Cost reduction;Decision making;Emission control;Investments;Sales ;Stochastic control systems;Stochastic systems;Taxation;Carbon emissions;Carbon emissions reductions;Carbon reduction;Carbon tax regulation;Carbon taxes;Chain management;Emission taxes;Loss-aversion;Reduction technologies;Technology investments
    • Zhao, Han;Yan, Lixun;Liao, Yu
    • 《44th Chinese Control Conference, CCC 2025》
    • 2025年
    • July 28, 2025 - July 30, 2025
    • Chongqing, China
    • 会议

    This paper studies one risk-neutral supplier and one loss-averse retailer supply chain with reference dependence under carbon emissions tax regulation, where stochastic demand is sensitive to carbon emissions. The optimal decision-making for both supplier and retailer in the supply chain are derived. Specifically, the loss-averse retailer's optimal ordering policy is obscure, and affected by the loss aversion, carbon tax, contract parameters, random demand and yield. It is shown that the optimal ordering quantity of the retailer decreases in the carbon emissions tax and loss aversion level. Further, The risk-neutral supplier's optimal carbon reduction technology investment level is affected by costs for carbon emission reduction, cost-sharing ratios, carbon emissions tax, contract parameters and random demand. It is demonstrated that the higher the retailer's sharing ratio of carbon emission reduction costs, the higher the supplier's optimal carbon reduction technology investment level. In addition, this paper shows that the call option contract improves the supply chain's the performance and decreases invalid carbon emissions. Finally, it is shown that the joint carbon emission reduction and coordination of the supply chain need to meet two conditions, which related to the demand distribution and carbon reduction technology investment level. © 2025 Technical Committee on Control Theory, Chinese Association of Automation.

    ...
  • 排序方式:
  • 1
  • /