Abstract
The characteristics of the path in length, smoothness, and safety are critical to the behavior of autonomous mobile robots. However, in searching-based global path planning, the potential heading of the robot is artificially restricted, so the resulting path typically consists of several polylines containing undesirable sharp bends, which sacrifices optimality and practicality in continuous space. To improve the optimality, we propose a Forward Search Optimization (FSO) algorithm to shorten the path planned by searching-based global algorithms such as Dijkstra, A*, D*, and D* Lite. To improve the practicability, we develop a Subgoal-based Hybrid Path Planning (SHPP) approach to smooth the path and keep it safe and comfortable from obstacles. Rather than formulating the smoothing process as a mathematical optimization problem, the SHPP approach combines subgoals and local algorithms (i.e., fuzzy inference system and artificial potential field) to meet the safety and smoothness requirements, which is essential for the practical application of the algorithm. Moreover, the guidance of subgoals solves the local minimum and goal unreachable problems of local path planning algorithms in complex environments. The simulation results show that the FSO and SHPP algorithms generate a short, smooth, and safe path.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call