Informed-TRRT*: An improved sampling-based path planning algorithm.

Path planning involves finding the optimal path in a known environment, allowing robots or vehicles to reach the destination as quickly and safely as possible from the starting point. In this domain, the RRT* (Rapidly-exploring Random Tree Star) algorithm has emerged as an advanced path planning approach, in handling challenges in complex environments. However, due to the slow convergence rate of RRT*, and the significant uncertainty in the generated paths, achieving stability in generating optimal paths within a fixed number of iterations poses challenges for the RRT* algorithm. This paper develops a novel algorithm, Informed-TRRT*, which combines the Theta* algorithm with the Informed-RRT* algorithm. The algorithm proposed in this paper rapidly obtains a short initial solution using the Theta* algorithm, narrows the sampling region based on this initial solution to improve sampling efficiency, and incorporates pruning algorithms to further enhance path optimality. Through comparative simulation experiments in environments of increasing complexity, the proposed algorithm generates paths with lengths optimized by 1.23% and 3.56% compared to the Theta* algorithm and Informed-RRT* algorithm, respectively, in environments with gradually increasing complexity; 19.53% and 8.76%; 6.45% and 9.07%. [ABSTRACT FROM AUTHOR]