Path Smoothing Algorithm Using Thin-Plate Spline

Abstract

Generating smooth trajectories is vital for achieving a stable and efficient movement for mobile robots. Most robot planners generate paths with many dispensable and sharp turns that may be infeasible to the robot's kinodynamic constraints, so a post-smoothing process is needed. We propose a smoothing approach that deploys thin-plate spline and genetic algorithm to deform smoothed trajectories. The simulation shows that the proposed approach improves the quality of the smoothed path in terms of path length and maximum curvature.