UUV Path Planning Method Based on Improved RRT*-APF Algorithm

The path planning algorithm for the combination of rapidly-exploring random tree star and artificial potential field (RRT<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">*</sup>-APF) has many redundant sampling points, large inflection points and corners, and path oscillations. An improved RRT*-APF algorithm is therefore proposed. The improved algorithm introduces a Bayesian network (BN)-based sampling strategy, which guides the random tree to explore path that meet the navigation constraints of unmanned underwater vehicle (UUV). Additionally, a goal-biased strategy accelerates the convergence of the random tree towards the goal endpoint. A gravitational dynamic adjustment factor is introduced to dynamically change the influence of gravitational force on the goal endpoint exploration. Simulation experiments and results show that the algorithm has obvious advantages in terms of path length, exploration time and number of iteration convergence.