This study presents an enhanced motion planning and trajectory optimization algorithm for Rapidly-exploring Random Trees (RRT), addressing the issues of excessive randomness, unnecessary path segments, and discontinuities in the trajectory that are evident in conventional RRT path planning algorithms. Initially, a dynamic sampling strategy was developed to minimize the search's randomness by managing the generation location of random points. Subsequently, the artificial potential field method was incorporated into the RRT algorithm to enable nodes to account for the influence of obstacles and target points during expansion; thus, mitigating the search's aimlessness. Redundant sections were then eliminated, and the path distance was reduced, all the while maintaining a safe distance from obstacles. Finally, by employing the minimum snap method in conjunction with flight corridors and time reallocation, the trajectory was rendered smooth and continuous, ensuring that physical quantities remained within acceptable limits without encountering obstacles. The comparative simulation results demonstrate that the proposed algorithm markedly decreases the number of nodes extended during the path search and reduces the search path length in comparison to the traditional RRT.