A Software Architecture for the Dynamic Path Planning of an Autonomous Racecar at the Limits of Handling

Abstract

Based on a software architecture for autonomous driving presented and tested in an autonomous level-5 race-car in 2018 this paper describes in detail the evolutionary enhancement of this software architecture. The architecture combines the autonomous software layers perception, planning and control, which were modularized in the core software. The focus of this paper is the detailed description of how we enhanced the software with a module for an object list creation, a module for the behavioral planning and a module for the creation of dynamic trajectories. These enhancements allow the car to overtake other cars and static obstacles autonomously when driving on the race track. Furthermore, we present with a high novelty value the software module for a vehicle performance maximization, which consists of a control performance assessment and a friction estimation. The software architecture displayed in this paper will be tested and evaluated in the Roborace Season Alpha in 2019.