Path Optimization Based on Time-Optimal Velocity Profile Generation
Ákos Nagy · 2025
Mathematical optimization-based techniques are becoming increasingly important in the rapidly developing field of robotics. Optimization methods offer unprecedented opportunities, from control systems and motion planning to high-level tasks. The presented method starts from an initial path provided by an existing geometrical path planner. The initial path is not yet suitable for the robot, as it does not consider certain limitations of the robot or may result in too slow traversal (it would not be time-optimal). The presented method modifies this initial path so that the resulting time-optimal trajectory meets the kinematic and possible dynamic constraints of the robotic system. The path optimization approach is validated using a car-like robot in a simulated environment. Due to the applied second-order Lagrange dynamics, the method can also be applied to other types of robots.