主讲人：Dong Ngoduy

邀请人：邹亚杰 副教授

时间：2025年10月15日下午13：30（Beijing Time）

地点：通达馆103会议室

主讲人简介：

Prof. Dong Ngoduy obtained his PhD in Traffic Flow Theory and Simulation from Delft University of Technology, Netherlands, in 2006. Since 2020, he has been serving as the Head of the Transportation Engineering discipline in the Department of Civil Engineering at Monash University, Australia. Previously, he was a Professor of Transportation Engineering at the University of Canterbury, New Zealand (2017–2020), and the head of the Spatial Modeling and Dynamics (SMAD) research group at the University of Leeds, UK (2016–2017).

Since 2022, Prof. Ngoduy has been serving as an Associate Editor for IET Intelligent Transport Systems, an Associate Editor for Transportmetrica Part A: Transport Science since 2016, and an Associate Editor for Transportmetrica Part B: Transport Dynamics since 2012. He will serve as a Guest Editor for Transportation Research Part C in 2025. He has also served as a panel member for the UK Research Council.

Prof. Ngoduy has achieved outstanding research results, receiving the UK Research Council Senior Research Fellowship (2011–2016) and being continuously listed among the top 2% of global scientists since 2020. His research interests include Connected and Autonomous Vehicles (CAVs), traffic flow theory and simulation, transport digital twins, and urban traffic network optimization. He has published over 100 academic papers, primarily in Transportation Research Parts A, B, C, E and the IEEE Transactions series.

主讲内容简介：

Traffic stability analysis is an important theoretical tool for revealing traffic flow states. Such analysis requires car-following models that can accurately reflect real traffic conditions. However, existing car-following models typically idealize driving dynamics, neglecting stochasticity and delay effects, which can lead to biased results. Although recent studies have analyzed the stability of either stochastic systems or delayed systems separately, the coupled effects of both—closer to real traffic situations—remain rarely explored.

This study aims to fill this research gap by introducing additive and multiplicative noise based on the Wiener process and multi-class delay factors into car-following models, and by deriving the corresponding stability conditions using the Lyapunov method. Consequently, we extend stability analysis from traditional delay-free stochastic systems to delayed systems, representing a fundamental methodological shift—from stochastic differential equations (SDEs) to stochastic delay differential equations (SDDEs).

Furthermore, unlike previous studies that primarily focus on long-wave stability, this research applies the Euler method to SDDEs to examine stability conditions across the full wavelength spectrum, enabling systematic comparisons between long-wave and short-wave stability. To further capture time-correlated acceleration noise in driving behavior, we extend the notion of stochasticity by introducing persistent estimation errors, and derive their stability conditions using numerical methods based on linear matrix inequalities (LMIs).

Finally, we validate the theoretical criteria through numerical experiments and simulations, and, based on parameter settings of stochastic delay OVM, FVDM, and IDM models, explore mechanisms of speed oscillations and traffic safety risks. The results show that both stochasticity and delay contribute to system instability: stochastic noise primarily affects the low-speed, short-wave region, whereas delay induces instability across the full wavelength range. Additionally, driver estimation errors weaken mean-square stability—the longer the error persists, the greater its negative impact on system stability, although the marginal effect gradually diminishes.

The innovation of this study lies in extending traffic stability analysis to stochastic delayed systems and providing a general analytical framework that can be applied to other types of delays and control schemes. The corresponding theoretical derivations and numerical analyses are expected to provide a foundation for establishing more realistic traffic stability conditions and reveal underlying mechanisms of driving behavior.

欢迎各位老师、同学参加同济高等讲堂！

交通学院研究生会
交通学院青年教师沙龙