Grey Wolf Optimizer for UAV Path Planning Based on Differential Evolution Mutation Strategy
Feiyang Lv, Likui Wang · 2024
As the application environments for drones become increasingly complex, there is a need for an efficient algorithm to handle drone path planning problems. However, the results of many algorithms may be infeasible or inefficient, especially when facing three-dimensional complex flight environments. This paper proposes a combination of the Grey Wolf Optimizer (GWO) and Differential Evolution (DE) algorithms in the mutation phase, which maintains exploratory capabilities while promoting exploitation, to address UAV path planning problems. In GWO-DE algorithms, the GWO and DE algorithms are well-coordinated, adding a balance between exploitation and exploration. Additionally, chaotic initialization is employed to alleviate the issue of overly concentrated initial populations. The parameters are handled nonlinearly, and different alpha wolves are employed at different stages to improve the efficiency of early-stage development and later-stage exploration. In the three-dimensional simulation of UAVs, the proposed GWO-DE outperforms GWO and DE. The UAV path lengths generated by GWO-DE are shorter than its competitors.