A Heuristic Virtual Tube Generation Method Based on GAN
The unmanned aerial vehicle(UAV)swarms control method based on virtual tube combines the advantages of trajectory planning algorithm and vector field algorithm to transform the complex environment traverse problem of UAV swarms into a centralized virtual tube planning and distributed vector field control problem so that the difficult traverse problems in complex environment of UAV swarms can be solved.Generative Adversarial Networks(GANs)is applied into virtual tube generation.Firstly,a GAN model that can generate regions of interest in a random environment is designed,and the model is used to generate guidance regions to quickly obtain generation curves,thereby the rapid construction of virtual tube is achieved.The experimental results show that the proposed GAN model has stronger robustness,and the virtual tube generation algorithm based on the region of interest can greatly accelerates the convergence speed under the condition of ensuring the traverse performance comparing to the traditional methods.
generative adversarial networkvirtual tubetrajectory planning,swarm control
国家科技期刊平台
NSTL
万方数据
