论坛主席:伍国华 中南大学 7月10日,9:00-12:00
腾讯会议号:252-900-387
内容介绍
报告1：Machine learning techniques for data-driven modeling of intelligent vehicles and transportation systems
专家介绍：徐昕，国防科技大学教授，博士生导师。主要从事智能无人系统的自主控制与机器学习等方面研究，获国家自然科学二等奖1项、湖南省自然科学一等奖2项，湖南省科技创新团队奖1项。主持国家自然科学基金杰出青年项目和重点项目、国家重点研发计划项目课题、973项目课题、装备预研项目等20余项。任中国自动化学会自适应动态规划与强化学习专业委员会副主任、平行控制与管理专业委员会副主任、机器人智能专业委员会顾问委员，中国指挥与控制学会无人系统专业委员会副主任。出版专著2部，发表SCI论文100余篇，代表性论文发表在 IEEE TNNLS, J. AI Research, J of Field Robotics, IEEE TSMC: Systems, IEEE TPAMI, IEEE TCST, IEEE TITS 等期刊。任 IEEE Transactions on SMC: Systems, Information Sciences, International Journal of Robotics and Automation、IET Cyber-systems and Robotics 等国际期刊的Associate Editor， CAAI Transactions on Intelligence Technology 副主编以及《控制理论与应用》编委。
报告2：Edge Computing for Cooperative VehicleInfrastructure System
报告摘要：Cooperative Vehicle Infrastructure System (CVIS) significantly orients the development of future Intelligent Transportation Systems (ITS) to improve traffic efficiency and vehicle safety, but it puts forward more requirements for the computing ability and security of the system. The emergence of edge computing makes it possible to implement a new generation of cooperative vehicle infrastructure technology. This talk presents general introductions of edge computing and related secure technologies and discloses their innovative application trends in CVIS. Moreover, we introduce the latest research and practice of our team on reliable computation offloading and secure data sharing in vehicular edge networks. We believe that secure and reliable applications of edge computing would bring broad development prospects for CV IS.
专家介绍：田大新，教授，博导，北京航空航天大学交通学院副院长，青年长江学者，国家自然科学基金委优秀青年基金获得者，牛顿高级学者，IEEE Senior Member，中国电子学会智能交通信息工程分会副主任、中国计算机学会智能汽车分会副主任、中国指挥与控制学会无人系统专委会副主任委员、中国人工智能和大数据百人会专家委员、中国智能交通协会“优秀青年专家”、中国交通教育优秀中青年教师；担任运动型多用途乘用车节能与智能化京冀联合实验室主任，车路协同与安全控制北京市重点实验室副主任，城市交通管理集成与优化技术公安部重点实验室车联网研究室主任。主持国家自然科学基金、国家重点研发计划等纵向科研课题11项，发表学术论文100余篇，出版专著7本、教材2本、译著2本；授权发明专利34项；获国家科学技术进步奖二等奖等科技奖13项，北京市教学成果奖一等奖2项；担任国际学术期刊《IEEE Transactions on Intelligent Vehicles》、《IEEE Internet of Things Journal》、《Complex System Modeling and Simulation》的Associate Editor。
报告3：:Micro-scale Searching Algorithms for Solving Two-Echelon Vehicle Routing Problems
报告摘要:The two-echelon vehicle routing problem (2E-VRP) is an important model for optimizing the logistics system of a new era. It is also an open problem in management science. This report describes the background, a basic mathematical model and research status of 2E-VRP. It focuses on introducing two intelligent optimization algorithms based on micro-scale searching for solving this problem. The first one is a graph-based fuzzy evolutionary algorithm. Based on the distance-target correlation analysis between customers, the fuzzy assignment graph of the population is constructed. Offspring are searched for in a smaller search space than the global search space. This method significantly improves efficiency of solving the 2E-VRP. Based on the assumption that the optimal solution of 2E-VRP satisfies the property of an embedded Hamiltonian graph, a heuristic algorithm is designed to conduct micro-scale searching in the subgraph structure that satisfies this property. This method greatly improves the solution accuracy of 2E-VRP. Finally, we show the applications and related industrialization cases of these research results in urban logistic planning and the business scheduling of large-scale ports.
专家介绍：黄翰，男，博士，华南理工大学软件学院教授、博士生导师，兼任国际学术期刊IEEE Transaction on Evolutionary Computation（IF: 11.554）副编、中国仿真学会智能仿真优化与调度专委会副主任、中国工业与应用数学学会大数据与人工智能专委会副秘书长、大数据与智能机器人教育部重点实验室副主任、广东省本科高校软件工程专业指导委员会主任委员、广东省数学会金融与人工智能专业委员会副主任委员、广东省财税大数据重点实验室信息技术咨询专家组组长、广东省运筹学会物流分会副会长、广东省计算机学会软件工程专业委员会秘书长、广东保险业咨询专家、广东省计算机学会第十一届区块链专委会副主任、广州工业与应用数学学会副主任、广东省大数据与计算广告工程技术研究中心技术委员会主任，CCF杰出会员和IEEE高级会员；主持国家级和省部级重大项目等共10多项课题，以第一作者或通讯作者身份在IEEE TCYB、IEEE TETC、IEEE TSE、IEEE TEVC、IEEE TIP、IEEE TFS和《中国科学》等专业学术期刊发表论文60多篇，代表作入选ESI；授权国家发明专利35项、美国发明专利4项；获广东省科技进步一等奖和广东省自然科学二等奖；长期致力于智能算法理论、应用与产业生态的研究，相关工作的代码可以从如下链接获得：https://www2.scut.edu.cn/huanghan/
报告4：Learning to Solve Vehicle Routing Problems
报告摘要: Vehicle routing problem (VRP) is the most widely studied problem in operations research (OR), which is always solved using heuristics with hand-crafted rules. In recent years, there is a growing trend towards exploiting deep (reinforcement) learning to automatically discover a heuristic or rule for solving VRPs. In this talk, I will first briefly introduce the construction type of neural heuristics, followed by the elaboration of improvement type. Then, I will present the challenges in this area and my personal thoughts on them.
专家介绍：曹志广，博士，现任新加坡科技研究局制造技术研究院研究员，此前在新加坡国立大学工业系统工程与管理系担任研究助理教授。近年来，曹志广博士的研究兴趣集中在基于学习的优化方法，尤其是利用深度(强化)学习求解各类组合优化问题如车辆路径问题，车间调度问题，装箱问题和整数规划问题等，属于人工智能和运筹学领域的热点研究方向。曹志广博士在该方向上的成果发表在NeurIPS, ICLR, AAAI, IJCAI等AI顶会以及IEEE会刊等，相关工作的代码可以从如下链接获得：https://zhiguangcaosg.github.io/publications/
报告5：Modeling and control technology for high maneuvering autonomous flight of fixed-wing UAV
报告摘要：Performing agile maneuvers autonomously remains a challenge for the fixed-wing unmanned aerial vehicles (UAVs) because the UAV is required to fly along complex trajectories while closing to its physical limits. This presentation provides an integrated modeling and control method to address this issue. First, an experienced pilot performs several sophisticated maneuvers manually. Then, instead of directly controlling the UAV to track the demonstrated trajectory, we split the teaching maneuvers into multiple segments. A novel dynamic motion primitive described by a unit dual quaternion (DQ-DMP) is then developed to encode and store those segments’ rotational and translational features. Besides, a robust connecting method for the learned DQ-DMPs is introduced, resulting in a smooth maneuvering trajectory that even the pilot has not taught yet. Additionally, a cascaded inversion control scheme with stable tracking performance is also developed. Our system is integrated into an autonomous fixed-wing drone and implements several complex maneuvers in the outdoor environment. The results of comparative experiments and outdoor flight reveal that our approach can learn, adapt and generalize maneuvers in real-world scenarios.
专家介绍：喻煌超，国防科技大学副研究员，加拿大阿尔伯塔大学博士，国防科技大学首批卓青人才，中国机器人大赛无人机挑战赛技术委员会委员，PHCI期刊编委，《无人系统技术》期刊青年编委会副主任，长期从事无人机设计和控制相关研究。作为负责人主持军科委重点项目等国家级项目6项、省部级项目1项；参与国家级项目14项。发表高水平论文39篇，其中VPP、RAL等SCI论文16篇（2篇ESI前1%），CDC、ICRA等顶会论文10篇，获国际会议最佳论文奖2次，授权/实审专利27项。
报告6：Data-driven vehicle routing methods and applications
报告摘要：This work focuses on data-driven vehicle routing model and algorithm, and introduces their applications in ground bus, bike sharing, car sharing, customized bus and other modes of transportation.
专家介绍：马红光，博士，北京化工大学经济管理学院讲师、硕士生导师。研究领域包括交通运输管理、运筹优化、大数据决策等，在IEEE Transactions on Fuzzy Systems、IEEE Transactions on Engineering Management、Information Sciences、Applied Soft Computing、系统工程学报等国内外重要期刊上发表学术论文20余篇，申请国内外专利11项（已授权4项），登记软件著作权6项。曾获中国运筹学会智能计算分会第十三届年会优秀论文特等奖、北京化工大学2019届优秀博士学位论文。担任中国优选法统筹法与经济数学研究会船海经济管理分会副秘书长、中国优选法统筹法与经济数学研究会智能决策与博弈分会理事；Information Sciences、Soft Computing、Scientific Programming等SCI期刊的客座编辑。
