报告题目: Neural Network-Based Adaptive Resilient Cruise Control

报告人：Xinzhi Liu（刘新芝），加拿大滑铁卢大学教授

邀请人：楼旭阳

报告时间：2025年6月3日（周二）上午 10：30-11：30

报告地点：伟德bv国际体育C317

报告摘要：In the realm of Intelligent Transportation Systems (ITSs), ensuring the safety and stability of connected automated vehicles (CAVs) is of paramount importance due to their susceptibility to vulnerabilities in interactions. The potential for system-wide disruption stemming from a cyber-attack on the leader underscores this need. Therefore, this talk introduces a nonlinear neural-Lyapunov-based adaptive resilient cruise control approach aimed at ensuring that all vehicles maintain safe tracking of the leader's profile, even in the presence of cyber-attacks and external disturbances. To achieve this, we employ an adaptive neural network to estimate the system's nonlinear characteristics. Subsequently, the control procedure is proposed, utilizing a virtual disturbance observer and Lyapunov theorem for stability analysis and adaptive laws to deal with nonlinearity, external disturbances, deception attacks, and singular control gain. Notably, our proposed approach eliminates the need for restrictive assumptions such as Lipschitz conditions on the nonlinear component and avoids the requirement for additional algorithms to switch between controllers in the event of a cyber-attack. It provides compelling evidence of system stability and the achievement of control objectives. Additionally, simulation and comparative results validate the theoretical analysis, highlighting the efficacy of the proposed methodology.

报告人简介：刘新芝现任加拿大滑铁卢大学应用数学系教授。1988年，在美国德克萨斯大学阿灵顿分校获博士学位。1988年-1990年，在加拿大阿尔伯塔大学作博士后。1990年，受聘于滑铁卢大学为助理教授，并于1994年被聘为该校终身教授。刘新芝教授入选国家高层次人才计划，其研究方向包括：混合动力系统、非线性系统控制及稳定性、复杂网络的同步、通讯网络模型、传染病模型等。已发表学术论文400余篇，学术专著6部，学术编著20余部。