An Improved Differential Evolution Algorithm Inspired by Nature for UAV Path Planning
Jin Fang, Yang Hui · 2025
Path planning is a key link in the process of UAV mission execution, which needs to calculate the optimal path on the basis of mission requirements and flight constraints. The traditional differential evolutionary algorithm has a rapid decline in population diversity as the number of evolutionary generations increases, and easily plunges into the local optimum and low optimization efficiency. To address these problems, this paper proposes a Improved differential evolution for UAV path planning. Our algorithm is divided into two phases: exploration and development. The exploration phase uses the improved DE algorithm to generate known solutions; the development phase uses the SBOA algorithm to introduce dynamic perturbation factors to balance the known and unknown solutions. Comparison results with SaDE, GWO and other algorithms ground show that the our algorithm performs well in terms of path length, fitness value, standard deviation, which has strong robustness and adaptability.