Trajectory planning for car-like vehicles: A modular approach

We consider the problem of trajectory planning with geometric constraints for a car-like vehicle. The vehicle is described by its dynamic model, considering such effects as lateral slipping and aerodynamic drag. We propose a modular solution, where three different problems are identified and solved by specific modules. The execution of the three modules can be orchestrated in different ways in order to produce efficient solutions to a variety of trajectory planning problems (e.g., obstacle avoidance, or overtake). As a specific example, we show how to generate the optimal lap on a specified racing track. The numeric examples provided in the paper are good witnesses of the effectiveness of our strategy.