学术报告题目：Online scheduling and distributed optimizations for cyber-physical systems
报告人：香港中文大学（深圳）唐晓莹副研究员
报告时间地点：2020年10月4日16：30 D125

Title: Online scheduling and distributed optimizations for cyber-physical systems

Abstract:

Cyber-physical systems, such as smart power grid and intelligent transportation network, are integrations of communication, computation, and control to physical processes. Currently, stable and efficient operations of cyber-physical systems are challenged by the large-scale deployment, complex correlations among interconnected entities, and the uncertainties from both user behaviors and other random factors. To overcome these challenges, my past research focuses on developing online, distributed, and computationally efficient methodologies for interconnected cyber-physical systems. In this talk, I will mainly introduce two research projects on energy systems and transportation networks. The tools developed in the two works, however, can find applications in the broader cyber-physical field.

The first project studies the classical problem of online scheduling of deadline-sensitive jobs and investigates its extension to electric vehicle (EV) charging scheduling in smart grid. The problem lies in the category of time-coupled online scheduling problems without availability of future information. We proposed an competitive online scheduling algorithm called ORCHARD for the EV charging scheduling problem, in order to minimize the energy cost and fulfil charging demands of all the EVs before their deadlines. We proved that ORCHARD achieves the best known competitive ratio 2.39 when the cost function is quadratic function of the load demand. This theoretic analysis can be extended to the case where the cost function is a general increasing convex function of the load demand. In order to enhance the scalability of ORCHARD, we proposed a low-complexity algorithm to replace the standard convex optimization method used in ORCHARD, where the optimality of the proposed algorithm is verified by KKT conditions.

The second project is integrated algorithm design by considering the coupling effects of both power systems and transportation networks. We formulated an optimization problem to jointly plan the route and schedule the charging/discharging for large-scale EVs. To overcome computational intractability of the optimization problem due to nonlinearity and nonconvexity of the underlying physical model, we proposed an approximate distributed algorithm that effectively reduces the computational complexity at the system operator and protect the EV users' privacy and autonomy. Besides, a proximal method was introduced to improve the convergence rate of the proposed algorithm. In addition, we provided an ADMM-based algorithm as a benchmark. Extensive simulations using real world data show that the proposed distributed algorithm can achieve near-optimal performance with relatively low computational complexity in different system set-ups.

Bio:

Xiaoying Tang (Used name: Wanrong Tang) is currently a tenure track assistant professor in the Chinese University of Hong Kong, Shenzhen. Before that, she worked as a Research Scientist/Postdoc at école polytechnique fédérale de Lausanne (EPFL) and the the Chinese University of HongKong (CUHK) respectively during 2016-2019. She received her Ph.D. degree from the Information Engineering Department, CUHK in 2016, and B.Eng. degree from the Yingcai Honors College, University of Electronic Science and Technology of China (UESTC) in 2011. Her current research interests include online scheduling, distributed algorithm design and optimizations in cyber-physical systems such as energy systems and intelligent transportation networks. She received the Best Paper Award of IEEE SmartGridComm 2013, the Student Travel Grant from the IEEE Communication Society in 2013, and the Talent Development Scholarship from Hong Kong Special Administrative Region Government Scholarship Fund in 2015. She serves as TPC Co-Chair of ICC 2019 Workshop-ICT4SG, PC of APCC'19, Globecom’19, PSCC’18, APCC'18, APCC'17, SmartGricComm’16, SmartGricComm’15, Reviewer of top journals such as IEEE Transactions on Smart Grid, IEEE Transactions on Power Systems, IEEE Transactions on Industrial Informatics, IEEE IoT Journal, etc.