Dynamic safety zone-based path planning for reduction of unwanted emergency braking

Sammanfattning: Autonomous vehicles are required to have higher standards of safety. To enforce this , different subsystems in the autonomous vehicle are categorised to different safety levels by the industry. The problem arises when these subsystems are required to interact with each other. The objective of this thesis is to build and study the interaction between two systems in an autonomous vehicle. The two systems here are the path planning and the emergency braking system (AEBS) which are assigned different safety levels. Model Predictive Control(MPC) is used to build the interaction between the path planning and the emergency braking system. The zone of activation of the emergency braking system is modelled as polygon and is called the safety zone for the autonomous vehicle. The shape of this polygon is a function of a decision variable of the MPC. With respect to the safety zone, Farkas’s lemma is used to further constrain the vehicle to stay within the boundary of the road and avoid static obstacles.Two scenarios are built and studied to see the implications of the interactions of these two different systems. To study this interaction, the simulations are carried out with and without the interaction of the safety zone with the path planning system. This helps us in visualizing and understanding the factors influencing the size and shape of the safety zone. Key results of this thesis are to demonstrate the unwanted activation of the AEBS and the adaptive shape changing nature of the safety zone as a part of the MPC. These results demonstrate the importance of the safety zone being a part of the MPC, to prevent unwanted activation of the AEBS. The analysis contained in this thesis provides a methodology by which these two different standard systems(AEBS and path planner) can be integrated in an autonomous vehicle.