Smart Control for Safer Roads: Development and Testing of Advanced Adaptive Cruise Control Systems

• Applicant: Professor Dr. Antoine Tordeux
• Subject Area: Traffic and Transport Systems, Intelligent and Automated Traffic
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 546728715

Stop-and-go waves, also known as phantom jams, are a common occurrence on highways that cause drivers to slow down and speed up again for no apparent reason. Stop-and-go waves are complex, emergent collective phenomena that can be observed every day on roads around the world. In addition to their scientific interest, stop-and-go dynamics impacts road user safety, their travel times, and the environment.Although the first studies date back to the early 1950s, achieving platoon stability and understanding the formation of stop-and-go dynamics in traffic remain challenging today. Current adaptive cruise control (ACC) systems available in the market exhibit unstable dynamics in experiments. In fact, many factors perturb the stability of uniform configurations and cause the platoon system to collectively oscillate. Stability requires the agents to be accurate and reactive. However, delays and inaccuracies are inherent in the control of the vehicle and the perception of the environment. The ACC controllers require dynamic compensation by using anticipation mechanisms and other filters to achieve collective stability conditions.The objective of this project is to develop, analyse and test safe and robust ACC systems that mitigate delay, lag, disturbances and stochastic noise in the dynamics. We incorporate selected linear and nonlinear car-following strategies in the upper level of advanced ACC systems and extend the control by using compensators and filters to attenuate latency, bias and noise operating at a lower control level. The delay induced by noise filters is also considered in a holistic modelling framework. The safety of the systems is assessed using strong linear stability criteria, including platoon (local) and string (collective) conditions. Both damped and overdamped stability dynamics are taken into account. Phase diagrams of selected nonlinear systems are simulated to identify phase separation, stability domains, corresponding critical parameter settings, and the operational limits of the developed ACC systems.Selected ACC systems are tested using a simulator and the open-source software CARLA. The focus is on safety-critical scenarios. Further experiments are conducted using down-scaled remote-controlled vehicles on a circular track. The objective is to test and evaluate the developed ACC systems in realistic situations by artificially playing with delays and perturbations in the dynamics. In addition, we aim to replicate and understand recent experiments involving ACC-equipped vehicles and to demonstrate the safety of the developed ACC systems.

• DFG Programme: Research Grants