经纬天地

针对无人机监测任务中路径规划效率低、易陷入局部最优的难题，提出一种改进萤火虫算法。改进算法基于传统萤火虫算法存在的初始种群分布不均、固定步长收敛慢等缺陷，引入Logistic混沌映射优化种群初始化，结合莱维飞行策略增强全局搜索能力，以提升算法的收敛速度与精度。通过构建含飞行速度、高度、航程约束的无人机监测路径模型，以某县域15个监测点的测绘任务为案例验证。结果表明：改进萤火虫算法（improved fireworks algorithm,IFA）迭代63次即收敛至最优解253.62 min，较传统萤火虫算法、粒子群算法（particle swarm optimization,PSO）和遗传算法（genetic algorithm,GA），其最优值分别提升2.3%、3.7%和6.3%,30次测试平均耗时265.84 min，稳定性优于对比算法。该方法可为复杂地形下无人机监测任务提供高精度路径规划支持，缩短规划时间，从而降低无人机的能源消耗，具有显著工程应用价值。

An improved firefly algorithm was proposed to address the challenges of the inefficient path planning and easily falling into local optima in drone monitoring missions. Based on the inequitable initial population distribution and slow convergence of fixed step lengths of traditional firefly algorithms, the improved algorithm introduced logistic chaos mapping to optimize population initialization, combined with Levi's flight strategy to enhance global search capabilities to improve the convergence speed and accuracy of the algorithm. By conducting a UAV monitoring path model that incorporates constraints on flight speed, altitude and range, a mapping mission involving 15 monitoring points within a county-level administrative region was used as a case study for validation.The results showed that the Improved Fireworks Algorithm(IFA) iterates 63 times converged to an optimal resolution of 253.62 min, compared to conventional fireworks algorithms, particle swarm optimization(PSO) and genetic algorithms(GA). Its optimal values were respectively improved by 2.3%, 3.7% and 6.3% and the average time of 30 tests was 265.84 min, with better stability than the comparison algorithm. This method can provide high-precision path planning supports for drone monitoring tasks in complex terrain, shorten planning time, thereby reduce drone energy consumption and it has significant engineering application value.