An improved black-winged kite optimization algorithm integrating sparrow behavior and multi-strategy optimization for UAV route planning

Abstract

This paper introduces an enhanced Black-winged Kite Optimization Algorithm (SMBKA) to address the issues of slow convergence and poor accuracy in UAV route planning within three-dimensional complex environments. SMBKA enhances planning performance by incorporating sparrow-inspired behavior and multi-strategy optimization mechanisms. Specifically, the proposed method utilizes Latin hypercube sampling to generate the initial population, thereby improving its distribution uniformity and diversity. Furthermore, an adaptive weight factor is adopted to achieve a dynamic balance between exploration and exploitation. Meanwhile, a dynamic opposition-based learning strategy is integrated to effectively improve overall search performance and avoid premature convergence. Finally, the scout search mechanism from the sparrow search algorithm is incorporated to enhance the algorithm’s adaptability to complex environments. In terms of performance verification, SMBKA demonstrates superior performance on the CEC2017 test functions, achieving the best average accuracy in 16 out of 29 test functions. In the application of UAV route planning, the average flight path length is shortened by 7.2% to 18.68%, demonstrating the potential of this algorithm in practical applications.