空间细胞机器人接管非合作目标时的交互式参数辨识方法

发布时间：2019-07-02 19:02

【摘要】：为应对日益复杂和多样的在轨服务需求,本文提出利用空间细胞机器人实施对非合作目标的接管控制.针对非合作目标接管控制中的惯性参数辨识问题,考虑到常用的递推最小二乘法依赖中心节点及计算量过大等不足,提出了一种适用于空间细胞机器人系统的交互式参数辨识算法,建立了质量、质心位置和惯性张量等惯性参数的辨识模型,利用细胞机器人之间的少量数据交互实现分布式和异步式的参数辨识.仿真结果表明,算法能够实现辨识结果的收敛性和一致性,辨识精度较高.最后分析了噪声、链路断开、拓扑结构和交互策略等因素对算法收敛性和一致性的影响,结果表明算法在噪声和链路断开等情况下仍能保证辨识结果的有效性.同时通过仿真分析验证了异步交互策略的优越性.
[Abstract]:In order to meet the increasingly complex and diverse demand for in-orbit services, this paper proposes to use space cell robots to take over non-cooperative targets. In order to solve the problem of inertial parameter identification in non-cooperative target takeover control, considering the shortcomings of the common recurrent least square method, such as relying on the center node and too much calculation, an interactive parameter identification algorithm for space cell robot system is proposed, and the identification model of inertia parameters such as mass, centroid position and inertia tensor is established. Distributed and asynchronous parameter identification is realized by using a small amount of data interaction between cell robots. The simulation results show that the algorithm can realize the convergence and consistency of the identification results, and the identification accuracy is high. Finally, the effects of noise, link disconnection, topology and interaction strategy on the convergence and consistency of the algorithm are analyzed. The results show that the algorithm can still ensure the effectiveness of the identification results in the case of noise and link disconnection. At the same time, the advantages of asynchronous interaction strategy are verified by simulation analysis.
【作者单位】：西北工业大学航天学院智能机器人研究中心;西北工业大学航天飞行动力学技术重点实验室;
【基金】：国家自然科学基金(11272256,61005062,60805034) 中央高校基本科研业务费(3102015BJ020)
【分类号】：TP242

【相似文献】