UAV Track Planning Based on Improved Sparrow Search Algorithm
Jie Yang, Fang Tuo, Yaoping Zeng · 2022 4th International Conference on Natural Language Processing (ICNLP) · 2022
UAV track planning is the planning of a safe and feasible track under environmental threats and its own constraints and is a prerequisite for UAVs to be able to perform a variety of tasks. To enable an unmanned aerial vehicle to quickly plan a safe and reliable track in complex environment, this paper presents an unmanned aerial vehicle track planning algorithm (ESSADE) based on the improved Sparrow Search algorithm. First the digital environment model of UAV flight is established, and the weighting of track length, flight height and turn angle is taken as the target function. Then, for the problem of sparrow search algorithm (SSA) initial population diversity is not rich, convergence rate is slow and easy to fall into the local optimal solution, elite reverse learning strategy is introduced in the SSA algorithm to widen the range of activities of the initial population and improve the diversity of the initial population. Then, the step size control parameters in SSA algorithm are dynamically adjusted to enhance the local search ability in the early stage and the global search ability in the later stage, so as to improve the accuracy of the algorithm. Finally, combined with the dynamic differential evolution algorithm, the sparrow population is divided into two sub populations, and the cross mutation operation is carried out respectively, and then the optimal sparrow is selected as a new species population to speed up the convergence speed, improve the ability to jump out of the local optimal solution and avoid premature phenomenon. The trajectory is finally smoothed by the B-sample difference. The simulation results show that the proposed ESSADE algorithm makes up for the shortcomings of SSA algorithm, and can better complete the track planning task in complex environment. The planned track is shorter, the height is the most stable, and the convergence speed is faster. The effectiveness and superiority of ESSADE algorithm are verified.