报告主题：Challenges in compliant robotic dynamics: advanced modeling and control mechanisms
报 告 人：Mihailo Lazarević 教授（贝尔格莱德大学）
报告时间：2025年10月29日 下午3：00
报告地点：江宁校区乐学楼1116
主办单位：力学与工程科学学院
欢迎广大师生参加！
报告简介：
Compliant (soft) robots represent a rapidly growing field of robotics that utilizes compliant and deformable materials to create systems capable of generating adaptive forces in unstructured environments. Due to their self-contained nature and inherent safety for human-robot interaction, these robots have a wide range of potential applications, including medical devices, industrial automation, machine inspection equipment, and environmental monitoring. However, they also present unique challenges in design and control, particularly regarding dynamic modeling, real-time control, non-linear material response behavior, and complex interactions with their environments. This presentation focuses on advanced compliant, extensible robotic systems based on a modular robotic platform that enables safe human-robot collaboration and scalable operation in dynamic medical and industrial applications. Some attention is given to the influence of discrete hybrid elements, including piezo-modified Kelvin-Voigt (PKV) and magnetorheological (MRD) viscoelastic models with fractional order derivatives, which are integrated into the system of motion equations through generalized forces. Furthermore, Iterative Learning Control (ILC) is a powerful method for enhancing the tracking performance of systems executing repetitive tasks. It is particularly well-suited for soft actuators (based on elastomer, hydrogel-elastomer, etc.) —such as soft robotic grippers and continuum robots—that exhibit nonlinear, hysteretic, and time-varying behavior. A self-tuning procedure is proposed for nonlinear robotic manipulators using model-free intelligent PID controllers optimized with Particle Swarm Optimization (PSO) algorithms, employing an ultra-local model formulation of the robots. Additionally, a novel control strategy is applied, combining an intelligent PD controller with a fractional-order iterative learning controller (ILC), where the fractional-order term in the ILC proves advantageous for tracking performance. Finally, the effectiveness of the proposed tuning approach is verified through simulation.
报告人简介：
Mihailo Lazarević is Full Professor at the University of Belgrade, Faculty of Mech. Eng. – Dep. of Mechanics as well chief of the Laboratory of Applied Mech. His current interests are related to application of fractional calculus to modelling and control of complex systems. So far, he published 1 International Monograph, 7 National Monographs, 8 Chapters in International Monographs, 27 articles in Leading International Journals, 28 articles in International Journals, 12 articles in National Journals,126 articles in the proceedings of International Meetings, Conferences and Symposiums, 1 Book and 2 Handbooks as collections of solutions and solved problems. According to Scopus, he has over 1546 citations and H index is 18 and according to Google Scholar he has 3067 citations and H index is 24. He is member of the IUTAM, and serves as reviewer for over 30 int. journals. From 2015-2019 he was President of the Serbian Society of Mechanics. He has been (co-)supervisor or a member of commission for over 35 diploma works, 3MSc theses and 12 Ph. D. dissertations. Six international research projects have been carried out since 2009. Also, Prof. Mihailo Lazarević is included in the ranked list Stanford list of 2% of the most cited scientists in the world for the year 2020, individually, and in the Stanford list of the 2% for 2022 and 2023, career-wise.
