【作者】 刘涛；

【导师】 张泽旭；

【作者基本信息】 哈尔滨工业大学 ， 航天工程（专业学位）， 2021， 硕士

【摘要】 在人工智能的推动下,针对无人系统的研究方兴未艾,无人机以其优越的灵活性和广泛的适用性在军民领域正发挥着愈来愈重要的作用。但囿于单架无人机的尺寸、能源及信息量等因素,难以满足多样化和复杂化的任务要求,尤其是GNSS拒止环境下信息不完整性使得系统面临诸多挑战。导航技术贯穿无人系统任务始终,而多无人机协同导航能够弥补单机导航性能的不足,通过机间信息的交互共享提高系统导航定位精度,增强鲁棒性。因此本文重点研究GNSS拒止环境下的无人机协同导航方法。首先,针对GNSS信号不可用但有地面参考点辅助测量的背景,建立基于卡尔曼滤波的多无人机协同导航状态估计理论框架,对多无人机的绝对测量和相对测量分别进行数学描述,推导验证基于高维卡尔曼滤波的集中式协同导航方法,通过多无人机仿真实验验证算法有效性,分析协同导航精度的影响因素。其次,在大范围GNSS拒止环境下,研究基于地形高程匹配的无人机自主导航方法,使用适于处理非线性问题的粒子滤波算法进行地形高程匹配,并针对匹配过程中的不确定性测量问题,提出融合无序测量更新的粒子滤波匹配算法,实现GNSS拒止环境下利用地形感知的无人机自主定位。最后,研究基于机间测距的无人机协同方法。仅通过机间测距信息,利用三角定位与间接平差法解算得到各无人机之间的相对位置信息,通过固定几何构型的约束实现对惯导信息的修正,然后基于无人机自定位与该协同方式实现多无人机自主协同导航。更多还原

【Abstract】 Driven by artificial intelligence,research on unmanned systems is on the rise,and drones are playing an increasingly important role in the civil-military field with their superior flexibility and wide application.However,due to the size,energy and amount of information of a single drone,it is difficult to meet the task requirements of diversification and complexity,especially in the GNSS-denied environment information incompleteness makes the system face many challenges.Navigation technology runs through unmanned system tasks,and multi-drone cooperative navigation can make up for the lack of single-machine navigation performance,improve the accuracy of system navigation and positioning through the interaction of information between UAVs,enhance robustness.Therefore,this paper focuses on the research of the GNSS-denied environment of the UAVs cooperative navigation method.First of all,in view of the background that GNSS signal is not available but has ground reference point assist measurement,the theoretical framework of multi-drone cooperative navigation state estimation based on Kalman filtering is established,the absolute measurement and relative measurement of multi-drone are mathematically described,and the centralized cooperative navigation method based on high dimension Kalman filtering is deduced and verified by multi-drone simulation experiments,and the influence factors of collaborative navigation accuracy are analyzed.Secondly,in the large-scale GNSS-denied environment,the autonomous navigation method of UAV based on terrain elevation matching is studied,the terrain elevation matching is carried out by using particle filtering algorithm suitable for dealing with nonlinear problems,and the particle filter matching algorithm of fusion disorder measurement update is proposed to realize the autonomous positioning of UAV using terrain perception in GNSS-denied environment.Finally,the cooperative method of drone based on inter-aircraft ranging is studied.Only through the ranging information,the relative position information between the drones is solved by triangular positioning and indirect parity method,the correction of inertial guidance information is realized by the constraints of fixed geometric configuration,and then multi-drone autonomous cooperative navigation is realized based on the self-positioning of the drone and the collaborative way.更多还原