A Particle Swarm Optimization Based Path Planning Method for Autonomous Systems in Unknown Terrain

Path planning of an autonomous system in unknown terrain is a challenging task. For a risk free and robust navigation, autonomous systems must utilize intelligence to determine the types of terrain and the traversability when optimizing its total cost (function). This paper presents a Particle Swarm Optimization based path planning for autonomous systems in unknown terrain environments. In this work, a new method is proposed toward terrain traversability analysis and estimation. Environmental data is gathered from sensors. Using this information, the proposed method identifies the terrain ahead and classifies them based on their traversability. Different weights are assigned against different types of terrain and these weights measure the characteristics of traversability on this terrain. The methodology autonomously plans a most traversable optimal path. Furthermore, this algorithm is capable to work in dynamic environments by avoiding collisions with obstacles. All simulations are carried out in MATLAB. Simulation results show the effectiveness and robustness of the proposed methodology.