Development of Safety control for Hidden Mode Hybrid Systems and Verification in the Multi-vehicle lab

In this thesis, we consider the safety control problem for Hidden Mode Hybrid Systems (HMHS), which are a special class of hybrid automata in which the mode is not available for control. For these systems, safety control is a problem with imperfect state information. We tackle this problem by introducing the notion of non-deterministic discrete information state and by then translating the problem to one with perfect state information. The perfect state information control problem is obtained by constructing a new hybrid automaton, whose discrete state is an estimate of the HMHS mode and is thus available for control. This problem is solved by computing the capture set and the least restrictive control map for the new hybrid automaton. Sufficient conditions for the termination of the algorithm that computes the capture set are provided. We show that the solved perfect state information control problem is equivalent to the original problem with imperfect state information under suitable assumptions on the original HMHS.;A multi-vehicle roundabout test-bed is developed that employs scaled vehicles that are designed to have longitudinal dynamical response similar to a full scale vehicle. The application of the proposed formal hybrid control approach to the collision avoidance problem between an autonomous vehicle and a human driven vehicle at a traffic intersection is experimentally illustrated in the multi-vehicle test-bed. We model the human driving behavior through a hybrid automaton, whose current mode is determined by the driver's decisions. On the autonomous vehicle, we implement formal methods for safety control, in which a mode estimator identifies in real time the current human driving behavior and uses this information to update a hybrid feedback map. The experimental results demonstrate that the solution proposed in this thesis is substantially less conservative than solutions employing worst-case design. Furthermore, they also demonstrate that, in structured tasks, human behavior can be reliably modeled and recognized for safety-critical closed loop control applications.