Modelling and Control of Test-Scenarios for Validation of Autonomous Driving Functions

Paper in proceeding, 2021

This paper presents a modelling and control framework for the coordination of leader-follower, multi-agent mobile systems, where the followers have to coordinate with a leading agent that is externally controlled. The problem setting is motivated by an experimental validation problem for a passenger vehicle (the leader), which has to be exposed to a traffic scenario by controlling a number of mobile test-targets (the followers) resembling the surrounding road users. The proposed control approach, which builds on reachability analysis and predictive control, allows to control the followers through the stages of the scenario, while satisfying the constraints imposed by the physical limitations of the followers, provided that the motion of the leader complies with certain bounds. Simulation results illustrate the proposed framework.