Title: Digital twin based robotic system for precision machining and assembly

Abstract: Due to the traditional off-line process planning cannot perform real-time control of manufacturing process through perception, it is difficult to fulfill high performance machining and assembly of major aerospace and energy equipment. Digital twin technology provides a novel solution for on-line optimization for machining and assembly automation.

The report is about digital twin based robotic systems for precision machining and assembly. First, the requirements and the challenges of precision machining and assembly technology for high-end equipment are analyzed. Next, the development and characteristics of digital twin is introduced, followed by several typical implementation modes. The framework and the key enabling technologies of digital twin systems are also discussed in detail. Then, the digital twin technology integrated with perception, modeling and simulation is proposed for precision machining and assembly. Finally, several digital twin based robotic systems are presented in high performance machining and assembly scenarios.

Bio: Dr. Haibin Yu is academician from Chinese Academy of Engineering，a full professor at the Shenyang Institute of Automation, Chinese Academy of Sciences, and he serves as the director of the State Key Laboratory of Robotics. He mainly engaged in research on industrial control networks and automation system technologies. He has hosted and participated in numerous national-level projects, including the National Natural Science Foundation of China (NSFC) Outstanding Young Foundation Project, the NSFC Key Project, the National High-Technology Research and Development Plan (863) Key Projects, and national major special projects. He has published over 100 papers in high-level domestic and international journals and at renowned international conferences. He has authored three academic monographs. His research achievements have won second prizes in the National Science and Technology Progress Awards three times and second prize in National Technological Invention Awards once. He is a recipient of the National Outstanding Young Foundation, and the Heliang He Li Science and Technology Progress Award. He also serves as a vice-president of the Chinese Association of Automation, and as a Fellow of the International Society of Automation (ISA).

Title: Collaborative Optimal Control and Safety Operation of Distributed Energy Interconnection Systems

Abstract:  Energy issues are the greatest challenge humanity faces in the 21st century. Therein, the distributed energy interconnection system is the product of the integration of traditional energy systems with modern information technology, forming a high-dimensional, strongly coupled, distributed dynamic system. In China, this clean, low-carbon, safe, and efficient distributed energy interconnection system with goals to peak carbon emissions by 2030 and achieve carbon neutrality by 2060. is willing to being built. However, optimizing collaborative decision-making and ensuring safe operation of distributed energy interconnection systems are urgent and significant scientific challenges at present. Thus, we have been engaged in research on optimization collaborative decision-making and safe operation of distributed energy interconnection systems for over 30 years. In optimization collaborative decision-making issue, the definition of iterative admissible control is firstly proposed, which becomes a common definition in the field of ADP. Secondly, the topological homeomorphism phenomenon between iterative optimal control and neural dynamics is discovered, and the principle of multi-dynamic homeomorphism compression is proposed. Thirdly, the basic properties of optimality (convergence and stability) of self-learning ADP are proved, and the optimal controller is made leapfrog from experience trial and error to fully automated design for the first time. Fourthly, a general framework for self-learning optimal control that can handle constraints is established through introducing a novel non-quadratic functional. Fifthly, we established a large-scale interconnection system's event-triggered self-learning optimal control mechanism, significantly reducing the system's computational and communication burden. This result is a universal proof of optimality in the ADP field so far. In order to apply into the distributed energy interconnection system, we reveal the intrinsic mechanisms linking coupled energy consumption and optimal cooperative indicators, establish an adaptive mechanism for updating approximate weight parameters, and propose an online distributed self-learning method for optimizing cooperative control. We establish a dual-impulse interval switching control strategy, forming a method for designing the stability domain of generalized eigenvalues through self-learning iteration. Based on this, the real-time performance and robustness of optimized cooperation under sudden working conditions is enhanced, achieving cost optimization including control costs. Furthermore, the exponential cooperative convergence with arbitrary precision within a specified time frame is ensured. Finally, the international challenges of optimizing cooperation among nonlinear multi-agents that current methods such as quadratic optimal, inverse optimal are overcame. Thus, the convergence speed regarding current and voltage is increased by 30% for power grid.In safe operation issue, the main problems are shown as follows: (1) The actual pipeline monitoring applies pressure data at both ends for leakage detection, but the traditional negative pressure wave analysis method cannot meet the needs of rapid alarms for weak pipeline leakage and accurate location of leakage points; (2) Foreign monopoly of pipeline inspection would make vital data of China leakage, leading to hidden dangers in energy information security, while foreign technical services provided by the submarine pipeline defect reconstruction accuracy is insufficient; (3) Existing land pipeline communication is unable to realize long-distance communication of submarine pipelines, and there is no available internal detector location technology, so it is urgent to develop a submarine pipeline safety monitoring system with independent intellectual property rights. Thus, we propose a chaotic synchronization-based fast detection for weak leakage. Firstly, a fuzzy hyperbolic tangent model is designed to model the pipeline fluid dynamics, and the chaotic dynamics of the pipeline fluid is applied to magnify the weak leakage signal. Secondly, data filtering and on-line self-tuning of localization parameters are constructed to reproduce the actual signal and accurately capture the location of sudden changes during leaks. Thirdly, adaptive pressure-transmission (P-T) dynamics model is constructed to locate leakage points.

Bio: Dr. Huaguang Zhang received the Ph.D. degree in thermal power engineering and automation from Southeast University, Nanjing, China, in 1991. He joined the Department of Automatic Control, Northeastern University, Shenyang, China, in 1992, as a Postdoctoral Fellow for two years, where he has been a Professor and the Head of the Institute of Electric Automation, School of Information Science and Engineering since 1994. He has authored and coauthored over 300 journal and conference papers, 6 monographs, and co-invented 90 patents. His main research interests are fuzzy control, stochastic system control, neural networks-based control, nonlinear control, and their applications. Prof. Zhang was awarded the Outstanding Youth Science Foundation Award from the National Natural Science Foundation Committee of China in 2003, the IEEE TRANSACTIONS ON NEURAL NETWORKS 2012 Outstanding Paper Award, and the A. P. Sage Best Transactions Paper Award 2015. He was named the Cheung Kong Scholar by the Education Ministry of China in 2005. He was an Associate Editor of the IEEE TRANSACTIONS ON FUZZY SYSTEMS from 2008 to 2013. He is an Associate Editor of Automatica, the IEEE TRANSACTIONS ON NEURAL NETWORKS, the IEEE TRANSACTIONS ON CYBERNETICS, and Neurocomputing. He is the E-Letter Chair of IEEE CIS Society and the Former Chair of the Adaptive Dynamic Programming and the Reinforcement Learning Technical Committee on the IEEE Computational Intelligence Society.

Title: PID Control with Intelligent Compensation

Abstract. Proportional-Integral-Derivative (PID) control has been widely used for regulating industrial processes. But its integral term can slow down the system's response (reduced bandwidth) and even lead to instability. This talk explores how to enhance PID control with intelligent compensation techniques. “PID Control with intelligent compensation” means a PID controller with an extra layer of intelligence. This 'intelligent' part utilizes machine learning techniques like neural networks or fuzzy logic. These data-driven compensators learn the system's non-linearities and disturbances without requiring a complex mathematical model. This is a significant advantage compared to traditional model-based compensation, which can be time-consuming and challenging to implement. The talk will delve into specific examples of designing neural and fuzzy logic compensators for PD control. These intelligent techniques can significantly reduce reliance on the integrator term in PID control. This not only improves the system's responsiveness but also allows for more flexibility in choosing the PID gains. We'll explore how Lyapunov stability analysis, a powerful mathematical tool, ensures the system remains stable even with reduced integrator action. Combining the strengths of traditional PID control with intelligent compensation can lead to faster response times, improved tracking accuracy, and greater robustness to disturbances in industrial processes.

Bio: Wen Yu received the B.S. degree in automation from Tsinghua University, Beijing, China in 1990 and the M.S. and Ph.D. degrees, both in automatic control, from Northeastern University, Shenyang, China, in 1992 and 1995, respectively. From 1995 to 1996, he served as a lecturer in the Department of Automatic Control at Northeastern University, Shenyang, China. Since 1996, he has been with CINVESTAV-IPN (National Polytechnic Institute), Mexico City, Mexico, where he is currently a professor with the Departamento de Control Automático. From 2002 to 2003, he held research positions with the Instituto Mexicano del Petróleo. He was a Senior Visiting Research Fellow with Queen’s University Belfast, Belfast, U.K., from 2006 to 2007, and a Visiting Professor with the University of California, Santa Cruz, from 2009 to 2010. He also holds a visiting professorship at Northeastern University in China from 2006.

He holds the distinguished position of Full Professor (Investigador Cinvestav 3F) at CINVESTAV-IPN in Mexico City, Mexico. He is a member of the Mexican Academy of Sciences. He currently has more than 500 publications, including more than 200 journal papers and 8 monographic books. He has supervised 38 PhD theses and 40 Master theses. His publications currently report more than 11,200 citations and his h-index is 52 according to Google Scholar. He is among the top 2% of the most-cited scientists in the world (Stanford/Elsevier, 2023). On Research.com's list of World's Best Scientists, in Electronics and Electrical Engineering as well as Computer Science, he holds the 6th position and 5th position in Mexico. He was the General Chair of the IEEE flagship annual meeting SSCI 2023. He serves as associate editors of IEEE Transactions on Cybernetics, IEEE Transactions on Neural Networks and Learning Systems, Neurocomputing and Journal of Intelligent and Fuzzy Systems.

Title: Intelligent Autonomous Systems: Dynamics, Control and Navigation

Abstract: Autonomous systems are broadly prevalent in many sectors, from manufacturing, agriculture, traffic management to medical industry. While tremendous progress has been made over the last decade in autonomous systems, many challenges still exist. This requires advances in many aspects of vehicle autonomy, ranging from design to control, perception, planning, coordination, and human interaction. The autonomous systems operating in complex, dynamic, and interactive environments require artificial intelligence that rapidly adapts to unpredictable situations. In this seminar, Prof. Shan will present some recent research outcomes on Intelligent Autonomous Systems (UAVs, UGVs and self-driving cars) from his research group, Spacecraft Dynamics, Control and Navigation Laboratory (SDCNLab) at York University. These topics include payload transportation using UAVs, UAV trajectory generation, game-theoretic decision making for autonomous driving vehicles.

Bio: Prof. Jinjun Shan is an internationally recognized expert in the areas of dynamics, control and navigation. He is a Full Professor of Space Engineering at the Department of Earth and Space Science and Engineering, York University. Prof. Shan received his Ph.D. degree from Harbin Institute of Technology, China, in 2002. His research progress is demonstrated through over 200 peer-reviewed journal and conference publications and 2 issued patents. Prof. Shan’s accomplishments in research and engineering education have seen him recognized with prestigious recognitions such as the Fellow of Canadian Academy of Engineering (CAE), the Fellow of Engineering Institute of Canada (EIC), the Fellow of American Astronautical Society (AAS), and a member of European Academy of  Sciences and Arts. He serves the profession as the Associate Editor for several field-leading journals including IEEE Transactions on Industrial Electronics, IEEE/ASME Transactions on Mechatronics, and the Journal of Franklin Institute, as well as numerous conference chairs.