Multi-UAV Collaborative Trajectory Planning for 3D Terrain Based on CS-GJO Algorithm

What is it about?

Existing solutions for collaborative trajectory planning using multiple UAVs suffer from issues such as low accuracy, instability, and slow convergence. To address the aforementioned issues, this paper introduces a new method for multiple unmanned aerial vehicle (UAV) 3D terrain cooperative trajectory planning based on the cuckoo search golden jackal optimization (CS-GJO) algorithm. A model for single UAV trajectory planning and a model for multi-UAV collaborative trajectory planning have been developed, and the problem of solving the models is restructured into an optimization problem. Building upon the original golden jackal optimization, the use of tent chaotic mapping aids in the generation of the golden jackal’s initial population, thereby promoting population diversity. Subsequently, the position update strategy of the cuckoo search algorithm is combined for purpose of update the position information of individual golden jackals, effectively preventing the algorithm from getting stuck in local minima. Finally, the corresponding nonlinear control parameter were developed. The new parameters expedite the decrease in the convergence factor during the pre-exploration stage, resulting in an improved overall search speed of the algorithm. Moreover, they attenuate the decrease in the convergence factor during the postexploration stage, thereby enhancing the algorithm’s global search. The experimental results demonstrate that the CS-GJO algorithm efficiently and accurately accomplishes multi-UAV cooperative trajectory planning in a 3D environment. Compared with other comparative algorithms, the CS-GJO algorithm also has better stability, higher optimization accuracy, and faster convergence speed.

Featured Image

Photo by Felipe Vieira on Unsplash

Photo by Felipe Vieira on Unsplash

Why is it important?

To address the issues of low accuracy, instability, and slow convergence speed in multi-UAV 3D collaborative trajectory planning solutions, this paper introduces the CS-GJO algorithm, which incorporates three key modifications. Firstly, the tent chaotic mapping is utilized for population initialization to enhance population diversity. Secondly, the CS algorithm is integrated with the GJO algorithm to enhance the GJO algorithm’s performance. Finally, specific nonlinear parameters are designed. In the early stages of the algorithm, these parameters help to accelerate convergence.

Perspectives