Collision-Free 4D Dynamic Path Planning for Multiple UAVs Based on Dynamic Priority RRT* and Artificial Potential Field

Abstract

In this paper, a four-dimensional (4D) dynamic cooperative path planning algorithm for multiple unmanned aerial vehicles (UAVs) is proposed, in which the cooperative time variables of UAVs, as well as conflict and threat avoidance, are considered. The algorithm proposed in this paper uses a hierarchical framework that is divided into a 4D cooperative planning layer and a local threat avoidance planning layer. In the cooperative planning layer, the proposed algorithm, named dynamic priority rapidly exploring random trees (DPRRT*), would be used for the 4D cooperative path planning of all UAVs involved in a given task. We first designed a heuristic prioritization strategy in the DPRRT* algorithm to rank all UAVs to improve the efficiency of cooperative planning. Then, the improved RRT* algorithm with the 4D coordination cost function was used to plan the 4D coordination path for each UAV. Whenever the environment changes dynamically (i.e., sudden static or moving threats), the proposed heuristic artificial potential field algorithm (HAPF) in the local threat avoidance planning layer is used to plan the local collision avoidance path. After completing local obstacle avoidance planning, the DPRRT* of the 4D cooperative planning layer is again called upon for path replanning to finally realize 4D cooperative path planning for all UAVs. The simulation and comparison experiments prove the feasibility, efficiency, and robustness of the proposed algorithm.

Adaptive Collision Avoidance for Multiple UAVs in Urban Environments

“…From Figure 14, it can be observed that due to the different locations and numbers of UAVs, the agent may determine different avoidance actions when facing collision risks, leading to different trajectories. DDQN is a typical value-based DRL algorithm [18,19], while APF utilizes the repulsive force of obstacles and the gravitational force of target to guide the UAV motion, and is widely used in research on collision avoidance [30,31]. DDQN requires that the agent action space be discrete, which is set to ∆𝜑 ∈ 3°, 0, 3° , ∆𝑍 ∈ 1 m, 0 m, 1 m , ∆𝑉 ∈ 2 m/s, 0 m/s, 2 m/s .…”

“…DDQN is a typical value-based DRL algorithm [18,19], while APF utilizes the repulsive force of obstacles and the gravitational force of target to guide the UAV motion, and is widely used in research on collision avoidance [30,31]. DDQN requires that the agent action space be discrete, which is set to ∆ϕ ∈ {−3 • , 0, 3 • }, ∆Z ∈ {−1 m, 0 m, 1 m}, ∆V ∈ {−2 m/s, 0 m/s, 2 m/s}.…”

Designing UAV Swarm Experiments: A Simulator Selection and Experiment Design Process

“…Potential field-based methods utilize attraction and repulsion forces to guide agents toward desired positions while avoiding collisions. Artificial potential functions [67] define a mathematical representation of the desired formation and drive the swarm toward it. Behavior-based methods focus on defining individual agent behaviors and interactions that collectively result in the desired formation.…”

Quad-Rotor Unmanned Aerial Vehicle Path Planning Based on the Target Bias Extension and Dynamic Step Size RRT* Algorithm

“…Path planning consists of planning an optimal path from an initial position to an end position, with quad-rotor UAVs flying in the presence of obstacles on the terrain and the required constraints being met at the same time [10]. At present, the most common algorithms for path planning include Dijkstra's algorithm [11], the A* algorithm [12][13][14], the Artificial Potential Field Method [15], the RRT (Rapid-exploring Random Trees) algorithm [16][17][18][19][20], and the RRT* algorithm [21][22][23][24][25], the last of which is progressively optimized for the RRT algorithm.…”