摘要
卫星姿态信息是影响整个姿态控制系统性能的重要因素。其精度不但取决于姿态测量系统硬件配置的性能与精度,还与所采用的姿态确定算法密切相关。星敏感器是一种矢量姿态敏感器,也是航天工程中精度最高的姿态敏感器。本文以星敏感器和陀螺作为基本配置组成的三轴稳定卫星姿态测量系统为对象,对卫星姿态确定的非线性滤波技术作了深入细致的研究。主要完成了以下几方面的工作:
首先较全面、系统地总结了各种姿态描述法、包括它们的定义、计算法则、运动学方程、换算关系及优缺点。基于四元数和修正罗德里格参数建立了完整的卫星姿态运动模型。
其次,采用四元数作为卫星姿态的描述,在测量模型为焦平面模型的情况下,推导出基于扩展卡尔曼滤波的星敏感器和陀螺组合的姿态确定算法。数值仿真表明扩展卡尔曼滤波算法在较小的初始误差下具有较好的性能,当初始误差较大时则不能保证收敛。
为了解决较大初始误差和非高斯分布的情况,重点研究了粒子滤波在卫星姿态和姿态角速度确定中的应用。分别针对有陀螺和无陀螺的情况,提出了用于姿态和姿态角速度确定的粒子滤波SIR算法。这种算法基于蒙特卡罗仿真,即用随机的粒子来近似表示状态矢量的概率分布。初始的姿态分布为均匀分布,采用罗德里格参数作为卫星的姿态描述。为了说明重采样方法对粒子滤波精度的影响,对3种不同的重采样策略进行比较。针对粒子滤波算法中出现的样贫问题,采用粒子粗化的方法来增加粒子的多样性。仿真中分别采用焦平面和星光矢量两种测量模型。仿真结果表明,粒子滤波SIR算法在使用粒子数为2000的情况下,虽然在小偏差下没有表现出更优越的性能,但是在大的初始误差和非高斯分布的情况下均有良好的收敛性能。而且,对于两种不同的姿态测量模型粒子滤波也同样具有很好的性能。同时,也可以发现,采用不同的重采样策略对最终的结果影响很小。
Satellite attitude is a critical piece of information in any space mission, and its accuracy is the key factor for the performance of the attitude control system (ACS). In general, the estimation accuracy not only lies on the performance of the hard-ware of the measurement, but also the attitude estimation algorithm. Star sensor is a kind of vector attitude sensor and is considered the most accurate absolute attitude sensor in aerospace engineering. In this thesis, the nonlinear filter methods are deeply studied for the attitude estimation from vector observations of the three-axis stabilized satellite attitude measurement system which composed of star sensor and gyro. The main contents of this dissertation consist of the following parts:
Firstly, a variety of attitude representation methods are generalized and analyzed, including their definitions, rules of calculation, kinematics, merits and disadvantages. Then the attitude kinematics and dynamic models about a rigid/flexible satellite are developed employing Quaternion and Euler-Angles representations.
Secondly, the Quarternion is chosen to describe the attitude state, and the measurement model is focal-plane model, the algorithm is derived under the measurement system composed of star sensor and gyro using Extended Kalman filter to deal with the vector observations from star sensors. The simulation indicated that under the small initial errors, the Extended Kalman filter has a good performance of convergence, but if the initial errors are large, the Extended Kalman filter failed to converge.
To solve the problem of large initial errors and non-Gaussian initial distribution, a new algorithm based on particle filter (SIR filter) for attitude and attitude rate determination is presented under the gyro and gyro-less mode respectively. The filter is based on Monte-Carlo simulation, and approximately represents the probability distribution of the state vector with random samples. The modified Rodrigues parameters are used to describe the attitude state. Use the uniform attitude probability distribution as the initial attitude distribution and a gradually decreasing measurement variance in the computation of the importance weights. To illustrate the effect of the resample process, three different resample strategies are used in the simulation. The particle impoverishment problem associated with the SIR filter is addressed by adopting a roughening process to the filter. Two measurement model focal-plane measurement model and LOS measurement model are used in simulation. Simulation results indicate that the particular particle filter, known as SIR filter, with as many as 2000 particles may not be as accuracy as EKF under small initial errors, but the convergence performance and accuracy is much better than EKF under large initial errors or non-Gaussian distribution. And for the two measurement model, the result showed that they also have almost the same accuracy and convergence. From the simulation, we can also found that different resample algorithms have the same effect on the accuracy of the particle filter.
引文
1杨大明.空间飞行器姿态控制系统.哈尔滨工业大学出版社, 2000
2周军.航天器控制原理.西北工业大学出版社, 2001.11:14~18
3李立宏.适用于星跟踪器的星图识别算法及飞行器姿态估计研究.哈尔滨工业大学工学博士学位论文. 2000:1~6
4 J. E. Pottr, W. E. Vander Velde. Optimum Mixing of Gyroscope and Star Tracker Data. Journal of Spacecraft.1967,5(5):536~540
5 N. Lovren, J. K. Pieper. Error analysis of direction cosines and quaternion parameters techniques for aircraft attitude determination. IEEE Transactions on Aerospace and Electronic System. 1998,34(3):983~989
6 C. C. Liebe. Accuracy Performance of Star Trackers- A Tutorial. IEEE Transactions on Aerospace and Electronic System.2002,38(2):587~599
7 G. Wahba. A Least Squares Estimate of Satellite Attitude, Problem 65.1 SIAM Review. July 1966: 385-386
8 M. D. Shuster, S. D. Oh. Three-Axis Attitude Determination from Vector Observations. Journal of Guidance and Control, 1981,4(1):70~77
9 M. D. Shuster. Approximate Algorithms for Fast Optimal Attitude Computation. AIAA 78~1249, AIAA Guidance and Control Conference, Palo Alto, CA, 1978
10 F. L. Markley. Attitude Determination Using Vector Observations and the Singular Value Decomposition. The Journal of the Astronautical Sciences,1988, 36(3): 245-258
11 F. L. Markley. Attitude Determination Using Vector Observations: A Fast Optimal Matrix Algorithm. The Journal of the Astronautical Sciences, 1993, 41(2): 261-280
12 D. Mortari. Euler-q Algorithm for Attitude Determination from Vector Observations. Journal of Guidance, Control, and Dynamics. 1998,21(2):328~334
13刘良栋等.卫星控制系统仿真技术.宇航出版社, 2003: 34-45
14 S. M. Joshi. On Attitude Estimation Schemes. IEEE Trans. on Aerospace and Electronic. 1978, for Fine-Pointing Control. 14(2): 258-265
15 E. J. Lefferts,F. L. Markley, M. D. Shuster. Kalman Filtering for Spacecraft Attitude Estimation. Journal of Guidance, Control and Dynamics. 1982, 5(5):417-429
16 T. W. Flatley, J. K. Forden, et al. On-board Attitude Determination and Control Algorithms for SAMPEX. Proc. of Flight Mechanics/Estimation Theory Symposium. NASA CP 3102, NASA Goddard Space Flight Center, Greenbelt, MD, 1990: 379-398
17 F. L. Markley, N. Berman, et al. Deterministic EKF-like Estimator for Spacecraft Attitude Estimation. American Control Conference. 1994,(1): 247-251
18 M. D. Shuster. The Quaternion in Kalman Filtering. AAS-93-553.1993:25-37
19 I. Y. Bar-itzhack, J. Oshman. Attitude Determination from Vector Observations: Quaternion Estimation. IEEE Trans. on AES. 1985, 21(1):128-135
20 M. D. Shuster. A Survey of Attitude Representations. The Journal of the Astronautical Sciences. 1993, 41(4): 437~517
21 S. Vathsal. Spacecraft Attitude Determination Using a Second-Order Nonlinear Filter. Journal of Guidance, Control, and Dynamics. 1987, 10(6): 559-566
22 S. Julier, J. K. Uhlmann. A General Method for Approximating Nonlinear Transformations of Probability Distributions. Technique Report. University of Oxford, U.K. 1994.
23 S. J. Julier The Scaled Unscented Transformation. Proceedings of the American Control Conference. Anchorage, Alaska USA, 2002: 4555-4559
24 M. S. Arulampalam, S. Maskell, N. Gordon, and T. Clapp. A Tutorial on Particle Filters for Online Nonlinear/Non-Gaussian Bayesian Tracking. IEEE Transactions on Signal Processing. 2002,50(2):174~175
25 N. Gordon, D. Salmond, et al. Novel Approach to Nonlinear/Non Gaussian Bayesian State Estimation. IEE Proc. of Radar and Signal Processing. 1993, 140(2): 107~113
26 J. S. Liu, R. Chen. Sequential Monte Carlo Methods for Dynamical System. Journal of the American Statistical Association. 1998. 93:103 2-1044
27 J. Carpenter, P. Clifford and P. Fearnhead. Improved Particle Filter for Nonlinear Problems. IEE Proceedings on Radar and Sonar Navigation, 1999,146(1): 2~7
28 A. Doucet, J. F. G. de Freitas, and N. Gordon. Sequential Monte Carlo Methods in Practice. Springer-Verlag. 2001
29 N. Gordon, D. J. Salmond, and C. Ewing. Bayesian State Estimation for Tracking and Guidance Using the Bootstrap Filter. Journal of Guidance, Control,and Dynamics.1995,18(6):1434-1443
30 F. Gustafsson, F. Gunnarsson, N. Bergman, and U. Forssell. Particle Filters for Positioning, Navigation and Tracking. IEEE Transactions on Signal Processing. 2002,50(2):425~437
31 K. Soohong and C. Joohwan. Satellilte Orbit Determination Using a Magnetormeter-Based Bootstrap Filter. Proceedings of the American Control Conference, Chicago, IL. 2000
32 Y. Oshman, A. Carmi. Spacecraft Attitude Estimation from Vector Observations Using a Fast Particle Filter. Proceedings of 14th AAS/AIAA Space Flight Mechanics Conference, Hawaii, Israel, 2004, AAS 04~14
33 Y. Cheng, J. Crassidis. Particle Filtering for Sequential Spacecraft Attitude Estimation. AIAA Guidance, Navigation, and Control Conference and Exhibit, Providence, Rhode Island, Aug. 16~19, 2004, AIAA-2004-5337
34章仁为.卫星轨道姿态动力学与控制.北京航空航天大学出版社, 1998:137~222
35黄圳圭.航天器姿态动力学.国防科技大学出版社, 1997:200~245
36 J. R. Wertz. Spacecraft Attitude Determination and Control. Boston, D Reidel, 1978:510~523
37 R. L. Farrenkopf. Analytic Steady- State Accuracy Solutions for Two Common Spacecraft Attitude Estimators. Journal of Guidance and Control. 1978,1(4):282~284
38 John L. Crassidis , John L. Junkins. Optimal Estimation of Dynamic Systems. Chapman & Hall/CRC, Boca Raton, FL, 2004,pp.419~430
39 A. Doucet, S. Godsill, and C. Andrieu. On sequential Monte Carlo sampling methods for Bayesian filtering. Statist. Computer, vol. 10, 2000,pp. 197-208
40 A. Doucet, J. de Freitas, N. Gordon. Sequential Monte Carlo Methods in Practice\. New-York:Springer-Verlag. 2001
41徐钟济.蒙特卡罗方法.上海科学技术出版社,1985
42裴鹿成,张孝泽.蒙特卡罗方法及其在粒了输运问题中的应用.科学出版社, 1980
43 Simon Godsill, Tim Clapp. Improvement strategies for Monte Carlo particle filters. Signal Processing Group, University of Cambridge.1998
44 A. Doucet. On sequential Monte Carlo methods for Bayesian filtering. Technicalreport, University of Cambridge, UK, Department of Engineering, 1998.
45 J. Carpenter, P. Clifford, and P. Fearnhead. An improved particle filter for non-linear problems. IEEE Proc., Radar Sonar Navigation.1999,146:2~7
46程杨.基于星敏感器的空间飞行器姿态和姿态角速度确定.哈尔滨工业大学博士论文. 2003
47姜雪原,马广富,胡庆雷.基于Marginalized粒子滤波的卫星姿态估计算法.控制与决策. 2007(1):39-44
48杨旭,程杨.粒子滤波在轨道确定中的应用.控制理论与应用. 2005(4):573-577