一、讲座安排
1. 讲座时间:2026年11月22日 9:00-18:30
2. 线下讲座地点:吉林大学南岭校区基础楼524
3. 线上会议入口:腾讯会议 603 697 020
二、日程安排
序号 |
报告题目 |
报告人 |
时间 |
地点 |
1 |
Predictive reduced-dimensional dynamics modeling with oblique projections |
莫焱芳 |
9:00-9:25 |
线上腾讯会议 |
2 |
Reconfigurable Distributed Antennas and Reflecting Surface (RDARS): A New Architecture for Wireless Communications |
马少丹 |
9:30-9:55 |
线上腾讯会议 |
3 |
面向协同驾驶的智能车辆运动行为建模与仿真测试关键技术 |
李永福 |
10:00-10:25 |
线上腾讯会议 |
4 |
面向灵巧操作的抓取生成方法 |
刘倩 |
10:30-10:55 |
线下报告 |
5 |
智能时代的汽车预测安全实时控制 |
王萍 |
13:00-13:25 |
线下报告 |
6 |
Stable and Robust Predictor for Linear Systems with a Long Input Delay |
Anton A. Pyrkin |
13:30-14:00 |
线上腾讯会议 |
7 |
Passivity Compensation: A Distributed Approach for Consensus Analysis in Heterogeneous Networks |
Lanlan Su |
18:00-18:30 |
线上腾讯会议 |
三、主讲内容
1. 报告题目:Predictive reduced-dimensional dynamics modeling with oblique projections
报告摘要:High-dimensional time series data are ubiquitous in modern industrial, cyber physical, and socio economic systems, yet their essential system dynamics often reside in a much lower-dimensional space. This talk focuses on predictive latent dynamics modeling, namely extracting low-dimensional dynamics that are genuinely predictive rather than merely temporally correlated. We introduce a probabilistic reduced-dimensional vector autoregressive (PredVAR) model that separates predictive latent dynamics from static or non predictive variations through oblique projections. By explicitly targeting predictability, the proposed approach enables parsimonious modeling and dynamically meaningful reconstruction and prediction in noisy, large scale systems. Its effectiveness is demonstrated through simulation studies and real industrial process data, highlighting the framework’s potential for predictive modeling, monitoring, and control with broad implications for industrial intelligence and smart infrastructure applications.
报告人简介:莫焱芳博士现任香港岭南大学助理教授,并担任工业数据分析理学硕士项目课程总监。她本科毕业于浙江大学控制科学与工程学院自动化专业(卓越工程师培养计划),随后获香港政府博士研究生奖学金(HKPFS),在香港科技大学电子及计算机工程学系取得博士学位,师从丘立教授。博士期间,她获海外研究奖学金,赴加州大学伯克利分校开展学术访问,合作导师为Pravin Varaiya教授。在完成博士学位后,莫博士先后于香港城市大学数据科学学院及香港数据科学研究院从事博士后研究,与秦泗钊教授、陈名华教授、李立帅教授合作,并曾任香港科技大学研究助理教授。她的研究方向包括工业人工智能,数据科学,在线决策、。相关成果已发表于IEEE TAC、Automatica、LAA、TASE、TII、Transportation Research Part C等国际知名期刊,并以第一作者身份获得IEEE INFOCOM 2021最佳海报奖。此外,她曾受邀担任Journal of Field Robotics特刊客座编辑,IEEE INFOCOM技术程序委员会成员,以及AMS Mathematical Reviews (MathSciNet) 特邀评论员。
2. 报告题目:Reconfigurable Distributed Antennas and Reflecting Surface (RDARS): A New Architecture for Wireless Communications
报告摘要:A new architecture "Reconfigurable Distributed Antennas and Reflecting Surface (RDARS) " will be introduced for future 6G wireless communications in this talk. Specifically, RDARS is a flexible combination of distributed antenna system (DAS) and reconfigurable intelligent surface (RIS). It inherits the low-cost and low-energy-consumption benefits of fully-passive RISs by default configuring all the elements as passive to perform the reflection mode. On the other hand, based on the design of the additional direct-through state, any element of the RDARS can be dynamically programmed to connect with the base station (BS) via fibers/wires and perform the connected mode as remote distributed antennas of the BS to transmit/receive signals. As such, this novel architecture exploits the benefits from both RIS and DAS with a controllable trade-off between the "reflection gain" and the "distribution gain" achieved via RDARS at the BS. Moreover, additional "selection gain" can be achieved from the reconfigurability of the operation modes for each element. Outage probability and ergodic achievable rate under maximum ratio combining (MRC) scheme at BS are analyzed in closed-forms to characterize the system behavior of the RDARS-aided system. Experimental results on a prototype of RDARS with 256 elements are also shown to demonstrate the superiority of the proposed RDARS. Particularly, the RDARS aided system with only one element operating in connected mode can achieve an additional 21% and 170% throughput improvement over DAS and RIS-aided systems, respectively. This confirms the high effectiveness and potential of the proposed RDARS for future 6G wireless systems. Finally, the RDARS-enabled new paradigm of integrated sensing and communications is briefly introduced with a practical demo to showcase the high flexibility and potential of RDARS.
报告人简介:Dr. Shaodan Ma is a Professor at the Department of Electronic and Communication Engineering, and Associate Director of the State Key Laboratory of Internet of Things for Smart City, University of Macau (UM). She received her Ph.D. degree in electrical and electronic engineering from The University of Hong Kong in 2006, and joined UM in 2011. Her research interests include array signal processing, transceiver design, localization, integrated sensing and communication, mmWave/THz communications, massive MIMO, and machine learning for communications.
She is an Executive Co-Chair of IEEE GLOBECOM 2026, and was a symposium co-chair for various conferences including IEEE VTC2024-Spring, IEEE ICC 2021, 2019 & 2016, IEEE GLOBECOM 2016, IEEE/CIC ICCC 2019, etc. She was an IEEE ComSoc Distinguished Lecturer in 2024-2025 and has served as an Editor for IEEE Transactions on Cognitive Communications and Networking (2025-present), IEEE Wireless Communications (2024-present), IEEE Communications Letters (2023), Journal of Communications and Information Networks (2021-present), IEEE Transactions on Wireless Communications (2018-2023), IEEE Transactions on Communications (2018-2023), and IEEE Wireless Communications Letters (2017-2022). She was the awardee of Macao Science and Technology Awards (3rd prize, Natural Science Award) in 2022.
3. 报告题目:面向协同驾驶的智能车辆运动行为建模与仿真测试关键技术
报告摘要:随着智能网联汽车向规模化应用发展,协同驾驶已成为提升交通安全与运行效率的重要技术路径,而高可信仿真与测试是支撑其研发验证和应用落地的重要基础。针对协同驾驶系统车辆、道路、通信与环境多要素耦合导致的建模困难、仿真可信度不足和测试效率受限等问题,本报告介绍团队在车辆行为建模与信息交互、协同驾驶仿真以及规模化智能测试方面的研究进展。报告围绕发展历程、研究现状、关键技术与典型应用,深入探讨智能车辆协同驾驶运动行为建模与仿真测试的研究思路、应用效果及未来发展趋势。
报告人简介:李永福,重庆邮电大学自动化学院院长,国家高层次青年人才、重庆市创新争先奖先进个人,“智能空地协同控制重庆市高校重点实验室”和“智能网联汽车与车路协同重庆市重点实验室”主任。担任IEEE T-IT、 T-Mech、T-ASE、T-IV和《自动化学报》等期刊高级副主编/编委。主持和承担国家重点研发计划等国家及省部级项目20余项。发表SCI论文100余篇,出版教材/专著4部。牵头荣获中国自动化学会青年科技奖/科技进步一等奖等8项。
4. 报告题目:面向灵巧操作的抓取生成方法
报告摘要:人类灵巧操作的能力是数百万年自然选择的产物——一只人手拥有超过20个自由度,能够以毫秒级的协调完成从拧开瓶盖到穿针引线的精密动作。长期以来,如何赋予机器人同等水平的灵巧操作能力,一直是机器人学最富挑战性的核心问题之一。早期研究主要聚焦于静态抓取位姿的生成,近年来随着生成式人工智能与物理仿真技术的快速发展,灵巧抓取研究正迎来从"静态位姿预测"向"动态操作执行"的范式转变。本报告重点介绍团队近期从静态位姿到连续动态操作的三项工作:基于细粒度接触引导的确定性抓取生成、基于解耦对齐表示的几何感知高效推理、以及基于反向轨迹策略的大规模动态数据合成与预训练,并探讨这些技术如何共同构建从接触到运动、从仿真到真实的灵巧操作技术链条。
报告人简介:刘倩,大连理工大学计算机科学与技术学院教授,博士生导师,主要研究方向为视触觉人机交互、多模态信号处理与通信。研究成果写入IEEE触觉编码标准(2024),获辽宁省技术发明一等奖(2021),国家级一流课程负责人(2025),曾任亚洲触觉大会程序委员会主席(2022)、中国触觉大会程序委员会主席(2026)。入选全球前2%顶尖科学家榜单(2024-2026)。
5. 报告题目:智能时代的汽车预测安全实时控制
报告摘要:在智能时代,预测控制方法凭借对未来状态的前瞻性优化能力,在汽车电控中得到广泛关注。汽车预测安全控制系统基于智能感知信息与实时状态反馈,协同主动安全控制目标,保证车辆稳定安全行驶,但优化问题求解的高计算复杂度也为控制系统快速实现带来了挑战。本报告首先介绍了预测控制方法基本特点,然后介绍预测安全控制系统的基本构成,最后通过案例从控制器设计、优化问题快速求解两个层面,展示了预测安全功能在纵侧向耦合运动控制、侧垂向耦合运动控制等方面的作用。
报告人简介:王萍,吉林大学控制科学与工程系教授,博导,国家级青年人才,长期从事预测控制理论及应用、复杂工况汽车主动安全控制等方面的研究工作。发表SCI论文30余篇,申请/授权发明专利20余项,主持国家自然科学基金面上项目2项,参与国家自然科学基金重大项目1项、科技部重点研发计划项目子课题1项;曾获吉林省自然科学奖二等奖1项(第一完成人)、吉林省科技进步奖二等奖1项(第二完成人);目前担任中国自动化学会预测控制与智能决策专业委员会委员以及具身智能专业委员会委员等。
6. 报告题目:Stable and Robust Predictor for Linear Systems with a Long Input Delay
报告摘要:The talk presents a new control algorithm for unstable linear systems with the long input delay. Unlike known analogues, the control law has been designed, which is the simplest to implement without requiring complex integration methods. At the same time, the problem of stabilization of a closed system is effectively solved, ensuring the boundedness of all state variables and the exponential stability of the equilibrium position.
报告人简介:Anton A. Pyrkin is the professor of the ITMO University, Saint Petersburg, Russia. He also serves as the head of the Department of Control and Robotics. His main research directions include nonlinear systems, adaptive observers, and robot control.
7. 报告题目:Passivity Compensation: A Distributed Approach for Consensus Analysis in Heterogeneous Networks
报告摘要:Consensus analysis in heterogeneous dynamical networks is an important research topic in distributed control, with applications in multi-agent systems, smart grids, and cooperative robotics. Classical passivity-based approaches to consensus analysis typically impose passivity requirements on individual agents. However, these requirements may be restrictive for heterogeneous networks containing agents with different dynamical characteristics, particularly when some agents exhibit a shortage of passivity. In this talk, we investigates a passivity-based approach to output consensus analysis in heterogeneous networks composed of non-identical agents coupled via nonlinear interactions, in the presence of measurement and/or communication noise. Focusing on agents that are input-feedforward passive (IFP), we first examine whether a shortage of passivity in some agents can be compensated by a passivity surplus in others, in the sense of preserving the passivity of the transformed open-loop system defined by the agent dynamics and network topology. We show that such compensation is only feasible when at most one agent lacks passivity, and we characterise how this deficit can be offset using the excess passivity within the group of agents. For general networks, we then investigate passivity compensation within the feedback interconnection by leveraging the passivity surplus in the coupling links to locally compensate for the lack of passivity in the adjacent agents. In particular, a distributed condition, expressed in terms of passivity indices and coupling gains, is derived to ensure output consensus of the interconnected network. The results provide a passivity-based framework for distributed consensus analysis and facilitate modular analysis and design of heterogeneous dynamical networks.
报告人简介:Dr Lanlan Su is a Lecturer (Assistant Professor) in Control Systems in the Department of Electrical and Electronic Engineering at the University of Manchester, UK. She received her B.Eng. degree in Electrical Engineering from Zhejiang University, China, in 2014, and her Ph.D. degree in Electrical and Electronic Engineering from the University of Hong Kong in 2018.
Following her Ph.D., she worked as a Postdoctoral Research Associate at the University of Notre Dame, USA, from 2018 to 2019. She subsequently held Lectureship positions at the University of Leicester (2019–2022) and the University of Sheffield (2022–2025), UK, before joining the University of Manchester in February 2025.
Her research interests include robust control, networked control systems, dynamical networks, distributed optimisation, and passivity and dissipativity theory. Her research has been published in leading international journals, including IEEE Transactions on Automatic Control, Automatica, IEEE Transactions on Control of Network Systems, and International Journal of Robust and Nonlinear Control.
Dr Su currently serves as an Associate Editor for the International Journal of Control and the European Control Conference. She has led research projects funded by the Royal Society and contributed to interdisciplinary research collaborations in control and systems engineering.