Invited Speakers

当前位置:首页  Invited Speakers  Session 10 Network Collaboration and ApplicationsSession 10 Network Collaboration and Applications

(10) Network Collaboration and Applications: Qingshan Liu, Souhteast University, China (Chair)

Shaofu Yang, Southeast University, China (Chair)

10.1. Speaker: Prof. Jianping He,  Shanghai Jiao Tong University, China

Talk Title: Learning and Attack via External Observation for Multi-Robot Systems

Abstract: In this talk, we propose a statistical learning framework for multi-robot systems to infer the underlying interaction relations and control rules by observing trajectories of interacting robots. A correlation algorithm adopted in Vector Auto-Regression (VAR) is designed and a transfer matrix estimator involving regression mechanism is proposed to infer the underlying topology (static or dynamic) of the robot system. Then, we design a learning-based intelligent attack scheme against the obstacle avoidance of formation control. With this scheme, the attacker can i) learn the detection area and the goal position of an agent through trial and observation; ii) regress the obstacle-avoidance mechanism using support vector regress (SVR) method. Experiments are provided to verify the feasibility and effectiveness of the proposed algorithms.

Biography: Prof. Jianping He (M’15) is currently an associate professor in the Department of Automation at Shanghai Jiao Tong University. He received the Ph.D. degree in control science and engineering from Zhejiang University, Hangzhou, China, in 2013, and had been a research fellow in the Department of Electrical and Computer Engineering at University of Victoria, Canada, from Dec. 2013 to Mar. 2017. His research interests mainly include the distributed learning, control and optimization, security and privacy in network systems. Dr. He serves as an Associate Editor for the IEEE Open Journal of Vehicular Technology and KSII Trans. Internet and Information Systems. He was also a Guest Editor of the IEEE TAC, International Journal of Robust and Nonlinear Control, etc. He was the winner of Outstanding Thesis Award, Chinese Association of Automation, 2015. He received the best paper award from IEEE WCSP’17, the best conference papers award from IEEE PESGM’17, and the finalist best student paper award from IEEE ICCA’17. He is the recipient of China National Recruitment Program of 1000 Talented Young Scholars.

10.2. Speaker: Prof. Huaqing Li,  Southwest University, China

Talk Title: Communication and Computation Trade-off Distributed Optimization Algorithm and Its Applications

Abstract: In distributed optimization, nodes cooperate to minimize a global objective function that is the sum (or average) of per-node local objective functions. Algorithms interleave local computations with communication among all the nodes.This talk will introduce the application backgrounds of distributed optimization, give an overview of the state-of-art distributed algorithms based on the structure of communication networks and the performance of convergence, and discuss how to implement distributed algorithms with fewer communications and computations while keep good convergence properties. Our recent work on an improved distributed primal-dual algorithm will be presented, where multiple primal updates are executed at each iteration. To avoid computational complexity, a local unbiased stochastic substitute of gradient is evaluated once per iteration and is used for all primal updates in that iteration at each node. This talk will focus on the trade-off between communication/computation and convergence performance of the proposed algorithm, and how the trade-off is involved in such implementations. Numerical tests on a classification problem using regularized logistic regression show the good performance and the advantage of the algorithm for solving large scale distributed statistical learning problems.

Biography: Prof. Huaqing Li is a currently a Professor at College of Electronic and Information Engineering, Southwest University, China. He received the Ph.D. degree in Computer Science and Technology from Chongqing University in 2013. He was a Postdoctoral Researcher at School of Electrical and Information Engineering, The University of Sydney from Sept. 2014 to Sept. 2015, and at School of Electrical and Electronic Engineering, Nanyang Technological University from Nov. 2015 to Nov. 2016. His main research interests include nonlinear dynamics and control, multi-agent systems, and distributed optimization. He was elected in Chongqing High-Level Personnel Special Support Program-Youth Top-notch Talent in 2018. He serves as a Regional Editor for Neural Computing & Applications and an Editorial Board Member for IEEE Access.

10.3. Speaker: Prof. Zhouhua Peng,  Dalian Maritime University, China

Talk Title: Path-guided distributed formation control of under-actuated autonomous surface vehicles

Abstract: Autonomous surface vehicles play an important role in studying, observing, protecting, and exploiting oceans. The collaboration capability is one of the key metrics to weight the intelligence of marine vehicles, and how to achieve a desired formation is central focus of current study among researchers. In this talk, some recent advances on coordinated control of marine vehicles are surveyed. Then, a distributed time-varying formation controller with capablity of collision avoidance and connectivity maintenance for a swarm of under-actuated autonomous surface vehicles guided by a parameterized path will be introduced. Finally, simulation and experiment results are presented to substantiate the efficacy of the proposed coordination method.

Biography: Prof. Zhouhua Peng received the B.E. degree in electrical engineering and automation, the M.E. degree in power electronics and power drives, and the Ph.D. degree in control theory and control engineering from Dalian Maritime University, Dalian, China, in 2005, 2008 and 2011, respectively. He held the postdoctoral position in the School of Control Science and Engineering, Dalian University of Technology, China, and the Hong Kong Scholar position in the Department of Computer Science, City University of Hong Kong, Hong Kong. Currently, he is a Professor with the School of Marine Electrical Engineering, Dalian Maritime University, China.

Dr. Peng was a recipient ofthe Science and Technology Award from Liaoning Province(2013,2017),the Youth Science and Technology Award from China Institute of Navigation, the Award of Hong Kong Scholar, the Natural Science Academic Achivement Award from Liaoning Province, the Science and Technology Achivement Award from Dempartment of Transportation, the Excellent Doctoral Dissertation Nomination Award from Liaoning Province, the Significant Academic Progress Award from Dalian Maritime University. He also won the honor of the BaiQianWan Talent in Liaoning Province, the Youth Science and Technology Talent in Ministry of Transport of China, the Science and Technology Star in Dalian, the Distinguised Young Teacher in Dalian, the Outstanding Youth in Dalian. Dr. Peng has authored/coauthored more than 160 papers in refereed journals and conferences. His research interests are guidance, control, and coordination of autonomous surface vehicles.

10.4. Speaker: Prof. Zai Yang,  Xi’an Jiaotong University, China

Talk Title: Positive Semidefiniteness of a Hadamard Product and Its Application in Array Signal Processing

Abstract: Two problems will be considered in this talk: 1) In array signal processing, how many antennas are sufficient to ensure the source resolvability? 2) In mathematics, under what conditions the Hadamard product of two positive semidefinite matrices is positive definite? We relate these two problems by providing a new Hadamard product perspective for the source resolvability. We give sufficient conditions for the second problem showing that both positive semidefinite matrices can be singular to ensure the positive definiteness of their Hadamard product. We apply this result to the first problem and show that fewer antennas than previously thought are sufficient to resolve the sources.

Biography: Prof. Zai Yang is currently in the School of Mathematics and Statistics, Xi’an Jiaotong University, China. He received the B.Sc. degree in mathematics and M.Sc. degree in applied mathematics from Sun Yat-sen (Zhongshan) University, China, in 2007 and 2009 respectively, and the Ph.D degree in electrical and electronic engineering from Nanyang Technological University (NTU), Singapore, in 2014. He has published over 30 peer-reviewed papers on the IEEE T-IT, IEEE T-SP and other journals and conferences. He is an IEEE Senior Member and serving on the editorial board of Signal Processing (Elsevier). His research interests include compressed sensing and optimization theory and their applications in signal and information processing, big data analytics, and machine learning.