Enhancing Passenger Comfort in Autonomous Vehicles Through Vehicle Handling Analysis and Optimization

Autonomous Vehicles (AVs) offer significant advantages in terms of traffic and fuel efficiency, accident prevention, and reduced travel time. 'Sense-plan-act' cycle of the AV has been designed and implemented to achieve a high level of active safety that considerably reduces the risk to passengers and pedestrians. However, within existing path planning and tracking algorithms, the subject of passenger comfort has received less attention compared to other AV topics. This paper is aimed at characterizing vehicle handling criteria and path profiles configuration in order to identify self-driving requirements for passenger comfort. The emphasis is given to the efficacy and usability of the classical handling analysis methods from a self-driving perspective. To these aims, we initially establish a framework which is defined by comfortable path profiles and a versatile handling model of an AV. Then, we introduce a directional path tracking unit which optimally implements AV's trajectories using a controller and a speed regulator. The path tracking unit includes a multi-objective optimizer enabling improvement the handling behavior of the AV. The integration of the proposed methodology into the state-of-the-art AVs' control system can lead to an enhanced comfort level.

Automated summary

This paper focuses on identifying the self-driving requirements for passenger comfort by characterizing vehicle handling criteria and path profiles configuration. Emphasis is placed on the efficacy and usability of classical handling analysis methods from a self-driving perspective.