无人机航迹规划算法综述OA
A review of unmanned aerial vehicle trajectory planning algorithms
为系统梳理无人机航迹规划领域的研究进展,本文首先对无人机航迹规划问题进行分析,依据算法原理对现有方法进行分类并介绍了其中常用算法的特点与应用;其次基于改进思路将近年研究归纳为基于算法自身局限性改进、结合环境表征改进以及基于多算法融合改进3类,最后指出航迹规划算法研究的难点与挑战以及现有研究的不足,然后在此基础上对未来发展趋势进行展望.研究结果表明,传统经典算法如A*算法、遗传算法、蚁群算法的改进已较成熟,而灰狼算法等新型智能算法以及与强化学习相结合的方法仍需深入研究.此外,当前研究主要针对单无人机场景,多机协同与复杂场景适应性仍显不足.需平衡环境建模的精度与效率,发展更贴合实际的建模方法并在此基础上优化规划算法.
To systematically review the progress in the field of unmanned aerial vehicle traiectory,this paper first analyzes the problem of UAV path planning and categorizes existing algorithms based on their algorithmic prin-ciples,highlighting the characteristics and applications of commonly used algorithms.Then,according to the im-provement ideas,the research of path planning algorithms in recent years is classified into the improvement based on the algorithm's own defects,the improvement based on environmental representation and the improvement based on multi-algorithm fusion.Finally,this paper points out the difficulties and challenges of path planning algo-rithm research and the shortcomings of existing research,and then looks forward to the future development trend on this basis.The research findings indicate that improvements to traditional algorithmsare relatively mature,while novel intelligent algorithms like grey wolf optimization and researches with reinforcement learning require further in-vestigation.What's more,current studies primarily focus on single-UAV scenarios,showing inadequacies in multi-UAV coordination and complex environment adaptability.Furthermore,it is essential to balance modeling accuracy with computational efficiency by developing more realistic modeling approaches to optimize trajectory planning algorithms.
王硕;李洋;赵蕴龙;刘春颜
南京航空航天大学 计算机科学与技术学院,江苏 南京 211106南京航空航天大学 计算机科学与技术学院,江苏 南京 211106||南京航空航天大学 无人机研究院,江苏 南京 211106南京航空航天大学 计算机科学与技术学院,江苏 南京 211106南京航空航天大学 计算机科学与技术学院,江苏 南京 211106
信息技术与安全科学
无人机航迹规划算法改进强化学习深度强化学习群体智能遗传算法算法融合
unmanned aerial vehicletrajectory planningalgorithm improvementreinforcement learningdeep reinforcement learningswarm intelligencegenetic algorithmalgorithm fusion
《哈尔滨工程大学学报》 2026 (3)
708-719,12
国家重点研发计划(2022ZD0115403).
评论