A novel exploration mechanism for addressing the issues of slow convergence speed, excessive redundancy and low path quality associated with the inherent randomness of Rapidly-exploring Random Tree (RRT)’s sampling approach is presented. First, the node exploration process of the RRT algorithm is modeled as a Markov Decision Process (MDP) by designing the action space and reward function. Subsequently, a novel node exploration mechanism based on RRT-Connect is developed by integrating environmental feedback information. Finally, the combination of Deep Q-Network (DQN) and RRT is achieved through the proposed DQN-RRT algorithm, which incorporates the structure and training method of DQN. Compared to the traditional RRT algorithm, the proposed algorithm balances planning autonomy, reduces search redundancy and efficient obstacle avoidance. Simulations are given to validate the performance optimization function of the proposed algorithm in RRT path planning.

Due to the complexity of three-dimensional (3D) map modeling, the A* algorithm is inefficient in planning 3D trajectories for unmanned aerial vehicles (UAVs). This paper designs a method for planning 3D trajectories using A* algorithm based on free space method to address this issue. The method translates the 3D trajectory planning problem into a two-dimensional (2D) trajectory planning problem through flight surface extraction, thereby enhancing the efficiency of trajectory planning. In addition, a trajectory optimization method considering safety parameters and offset costs is designed for the waypoints that fail to meet UAV restrictions. The optimization method consists of longitudinal trajectory altitude adjustment and horizontal trajectory smoothing to ensure the trajectory aligns with the UAV climb rate constraint, normal overload constraint and minimum turning radius constraint. Simulation results show that the designed method delivers higher efficiency in 3D trajectory planning and optimization than the conventional method, and the optimized trajectory adheres to the above constraints.