讲座题目：Data/system science-based analysis and design of nonlinear dynamical systems
主 讲 人：Yunpeng Zhu, Assistant Professor, Queen Mary University of London
讲座时间：2024年8月14日（周三）下午14：30时
讲座地点：江南大学理学院钱伟长楼201会议室
欢迎有兴趣的师生前来聆听！
理学院
2024年08月12日
讲座内容简介：
AI/machine learning places significant roles on data-driven modelling of complex systems. This is typically accomplished through the utilization of machine learning algorithms or system identification techniques, relying on either input and output data or solely output data from the concerned system. Many methodologies for data-driven modelling have been developed, including NN (Neural Networks), LSTM (Long-Short Term Memory) networks, etc. These contemporary methodologies have found extensive applications in tackling challenges in engineering systems design, control, condition monitoring, and characterization, primarily because of their impressive predictive and classification capabilities. However, applying a pure machine learning model in relevant engineering applications is just like a tightrope walking. The black-box nature of these models makes it challenging to place trust in them. In this talk, I will introduce the unique NARX (Nonlinear Autoregressive models with Exogenous Input) modelling and nonlinear frequency analysis methodologies I have developed since my PhD at the University of Sheffield. The research outcomes facilitate the application of physically interpretable machine learning in condition monitoring and engineering systems design. These involve (i) Condition monitoring of machining cutting tools, and (ii) Digital design of complex dynamic materials. The proposed analysis and design framework can be applied to resolve a wide range of dynamic problems in engineering practice.
主讲人简介：
Dr. Yunpeng Zhu is an Assistant Professor in the School of Engineering and Material Science at Queen Mary University of London. Dr. Zhu Holds a BSc (2013) and MSc (2015) in Mechanical Engineering from Northeastern University (China). He received a PhD in Automatic Control and Systems Engineering in 2020 and started to work in 2018 as a Research Associate specializing in the theories and methods for complex systems analyses and design and relevant applications in mechanical engineering and advanced manufacturing. His PhD thesis was published by Springer-Nature in the selection of the very best PhD theses from around the world and across the physical sciences. Dr. Zhu has published over 50 journal papers on Automatica, IEEE Trans, MSSP, etc. on the topics of data-driven science, frequency analysis, nonlinear dynamics, systems design and condition monitoring.
