Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Сен. 17, 2024

With the wave of artificial intelligence sweeping world in recent years, UAVs is widely used various fields. UAV path planning has attracted much attention from scientists as an essential part work. In order to design efficient and reasonable 3D program, researchers have invented improved many algorithms. This paper proposes elite RIME algorithm for planning. First, we propose reverse learning population selection strategy based on piecewise mapping enhance diversity better exploration. Second, this a stochastic factor-controlled pool exploration so that difficult enter local optimum can explore global optimum. Then, hard frost puncture exploitation sine-cosine function find faster during process. Meanwhile, test performance proposed paper, compare it with 13 other intelligent optimization algorithms are classical popular nowadays 52 functions three sets, CEC2017, CEC2020, CEC2022, obtain competitive results. Finally, applied problem different terrain scenarios, ELRIME achieved good results all them. Especially 7-peak model, improves by factor two. 9-peak average value aspect also reduce cost 91 compared algorithm, more importantly, smallest fluctuation 30 runs, which among most stable 12-peak its stability significantly enhanced, terms worst-case cost, 340 RIME.

Язык: Английский