在摄影测量领域，提升测绘卫星姿态确定精度，保证摄影时刻影像姿态信息的完整性和准确性，是实现目标定位尤其是无控目标定位的前提。本文以星敏感器，陀螺敏感器测量数据为基础，结合卡尔曼滤波理论，研究基于星敏感器和陀螺的卫星联合定姿算法，完成提升测绘卫星姿态确定技术精度和效率的目标。本文的主要完成工作和创新点如下：

（1）针对姿态确定系统的非线性问题，本文分析了扩展卡尔曼滤波（extended kalman filter，EKF），平方根无迹卡尔曼滤波（squared root unscented kalman filter，SRUKF）处理系统非线性的原理，构建了基于EKF，SRUKF 的星敏感器陀螺联合定姿框架。针对EKF定姿精度较差，SRUKF计算效率较低的问题本文提出了简化的SRUKF算法，分析了各算法的定姿精度和时间复杂度。通过不同精度星敏感器和陀螺数据条件下的仿真实验，分析各算法的特点和优势，实验结果表明简化的SRUKF算法相比于SRUKF算法，简化算法效率提高了35%左右，且简化算法平均定姿精度和稳定性更好。

（2）针对星敏感器测量值存在低频误差的问题，本文分析了低频误差成因，使用傅里叶级数的形式对低频误差进行建模；研究了有色噪声条件下卡尔曼滤波的增广状态方程和量测方程，设计实现基于状态扩维的EKF算法，对姿态和低频误差系数一同进行估计。利用仿真姿态数据，验证了算法的效果并讨论了不同状态扩展维数对于定姿结果的影响。

（3）根据低频误差的观测特性，本文提出使用滤波过程的状态残差来对低频误差进行检测。针对扩维EKF的不足提出迭代扩维EKF算法来检测和补偿姿态低频误差。仿真实验表明，对于测量数据中含有低频误差的定姿场景，迭代扩维EKF联合定姿算法能取得更佳的定姿精度。

（4）为进一步提高姿态精度，本文设计实现了基于双向EKF的联合定姿算法，能减轻定姿结果受不良滤波初始值和EKF线性化误差影响。基于双向滤波定姿和神经网络误差补偿思想，结合径向基函数（radial  basis function，RBF）提出改进的双向滤波算法。使用高斯噪声和低频噪声条件下的星敏感器数据和陀螺仿真数据对算法进行验证，实验证实RBF神经网络改进的双向滤波算法可以提高姿态确定的精度。

为实现测绘卫星姿态确定和质量提升的目标，本文从速度和精度两个角度对现有算法进行了改进。用姿态敏感器数据开展相关实验，证明了本文算法的优越性。

In the field of photogrammetry, improving the accuracy of surveying and mapping satellite attitude determination and ensuring the integrity and accuracy of the image attitude information at the time of photography are the prerequisites for achieving target positioning, especially uncontrolled target positioning. Based on the measurement data of star sensor and gyro, combined with kalman filter theory, this thesis studies the satellite attitude estimation algorithm based on star sensor and gyro to complete the goal of improving the accuracy and efficiency of the satellite attitude determination algorithm. The main work and innovations of this thesis are as follows:
(1) Aiming at the non-linear problem of the attitude determination system, the principles of EKF, UKF, and SRUKF dealing with system non-linearity are analyzed, combined with attitude kinematics theory and sensor measurement model, and a combination of star sensor and gyro framework based on EKF and SRUKF is constructed. Aiming at the problem of poor EKF attitude determination accuracy and low calculation efficiency of SRUKF, a simplified SRUKF algorithm is proposed, and the attitude determination accuracy and time complexity of each algorithm are analyzed. Through simulation experiments with different precision star sensors and gyro data, we analyse the characteristics and advantages of each algorithm, the experimental results show that compared with the SRUKF algorithm, the simplified algorithm has an efficiency increase of about 35\%, and the simplified algorithm has better average attitude determination accuracy and stability.
(2) Aiming at the problem of low-frequency error in the measured value of the star sensor, the mechanism of low-frequency error is analyzed and the low-frequency error is modeled in the form of fourier series. The expanded state equation and measurement equation of kalman filter under colored noise conditions are studied. The EKF algorithm based on state expansion is designed and implemented, whose attitude and low-frequency error coefficients are estimated together. Using simulated posture data, the effect of the algorithm is verified and the influence of the extended dimensionality of different states on the result of the posture determination is discussed. 
(3) According to the observation characteristics of low frequency error, this paper proposes to use the state residuals of the filtering process to detect low-frequency errors. Aiming at the deficiencies of the extended dimension EKF, an iterative extended dimension EKF algorithm is proposed to detect and compensate the low-frequency error of the attitude. The simulation experiments show that for the scenes with low-frequency errors in the measurement data, the iterative expanded EKF joint attitude determination algorithm can achieve better attitude determination accuracy.
(4) To further improve the attitude accuracy, this paper designs and implements a joint attitude fixing algorithm based on bidirectional EKF, which can mitigate the influence of bad filtering initial values and EKF linearization errors on the attitude fixing results. Based on the idea of bi-directional filtering and neural network error compensation, combining radial basis function, an improved bi-directional filtering algorithm is proposed. The algorithm is verified by using star sensor data and gyro simulation data under Gaussian noise and low frequency noise conditions, and the experiments confirm that the improved bidirectional filtering algorithm of RBF neural network can improve the accuracy of attitude determination.
To achieve the goal of attitude determination and quality improvement of mapping satellites, this paper improves the existing algorithm from both speed and accuracy perspectives. The relevant experiments are carried out with attitude sensor data to prove the superiority of the algorithm in this thesis.

[1] 谭政, 相里斌, 吕群波, et al. CX-6(02)微纳卫星超分辨率成像[J]. 遥感学报, 2019, 023(2) : 196 – 204.
[2] C J W A B, B R W A, B X H A, et al. The on-orbit calibration of geometric parameters of the Tian-Hui 1 (TH-1) satellite[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2017, 124 : 144 – 151.
[3] 王任享, WANGREN-XIANG. 我国无地面控制点卫星摄影测量综述[J]. 海洋测绘, 2008, 28(5) : 1 – 8.
[4] 朱红, 刘维佳, 张爱兵. 光学遥感立体测绘技术综述及发展趋势[J]. 现代雷达, 2014(6) : 6 – 12.
[5] 林宝军. 基于星敏感器的姿态控制与自主定轨算法研究[J], 2013.
[6] BLUMENTHAL D J. Chip-scale optical gyros based on integrated ultra low loss waveguide coils and silicon photonic front ends[C] // 2016 IEEE Photonics Conference (IPC). 2016.
[7] WANG P, ZHANG Y C, QIANG W Y. Research on the Algorithm of On-Orbit Calibration Based on Gyro/Star-Sensor[C] // Machine Learning and Cybernetics, 2006 International Conference on. 2006.
[8] KALMAN R. A new approach to linear filtering and predicted problems[J]. Journal of Basic Engineering, 1960, 82.
[9] LEFFERTS E J, MARKLEY F L, SHUSTER M. Kalman Filtering for Spacecraft Attitude Estimation[J]. Journal of Guidance Control Dynamics, 1982, 5(5) : 536 – 542.
[10] R. S, BUCY,, et al. Digital synthesis of non-linear filters[J]. Automatica, 1971.
[11] JULIER S J. Unscented filtering and nonlinear estimation[J]. Proceedings of the IEEE, 2004, 92(3) : 401 – 422.
[12] ARASARATNAM I, HAYKIN S. Cubature Kalman Filters[J]. IEEE Transactions on Automatic Control, 2009, 54(6) : 1254 – 1269.
[13] 刘垒, 张路, 郑辛, et al. 星敏感器技术研究现状及发展趋势[J]. 红外与激光工程, 2007, 36(S2) : 529 – 533.
[14] SCHMIDT U, ELSTNER C, MICHEL K. ASTRO 15 Star Tracker Flight Experience and Further Improvements Towards the ASTRO APS Star Tracker[C] // Aiaa Guidance, Navigation & Control Conference & Exhibit. 2013.
[15] AVANESOV G A, KRASIKOV V A, NIKITIN A V, et al. Accuracy estimation for determining the orientation parameters of the coordinate system of the BOKZ-M astromeasuring device by the experimental data[J]. Cosmic Research, 2015, 53(4) : 272 – 285.
[16] KRUZHILOV I S, KUNIAEV V V, FEDOSEEV V I, et al. Flight experience of 329K star tracker[J]. Proceedings of SPIE - The International Society for Optical Engineering, 2013, 8889 : 20.
[17] 向明江, 信思博, 姜通, et al. 高分五号卫星星敏感器在轨测量精度评估[J]. 上海航天, 2019(S2).
[18] SHUSTER M D, OH S D. Three-axis attitude determination from vector observations[J]. J Guidance Control Dynam, 1981, 4(1) : 70 – 77.
[19] WAHBA G. A Least Square Estimate of Spacecraft Attitude[J]. Siam Review, 1965, 8(3).
[20] MARKLEY F L. Attitude Determination Using Vector Observations and the Singular Value Decomposition[J]. Journal of the Astronautical Sciences, 1987, 38(3).
[21] MARKLEY F. Attitude Determination from Vector Observations: A Fast Optimal Matrix Algorithm[J]. J of Astronautical Sciences, 1993, 41(2).
[22] MORTARI D. Euler-q Algorithm for Attitude Determination from Vector Observations[J]. Journal of Guidance Control & Dynamics, 1998.
[23] BOWEN, HOU, ZHANGMING, et al. Integrated Design and Accuracy Analysis of Star Sensor and Gyro on the Same Benchmark for Satellite Attitude Determination System[J]. IEEE/CAA Journal of Automatica Sinica, 2019, v.6(4) : 213 – 219.
[24] 矫媛媛. 最优估计理论及在高精度卫星姿态确定中的应用[D]. [S.l.] : 国防科学技术大学, 2012.
[25] RISFIC B, ARULAMPALAM S, GORDON N. Beyond the kalman filter - Book Review[J]. Aerospace & Electronic Systems Magazine IEEE, 2004, 19(7) : 37 – 38.
[26] KAU S, KUMAR K S P, GRANLEY G B. Attitude Determination Via Nonlinear Filtering[J]. IEEE Transactions on Aerospace & Electronic Systems, 1969, AES-5(6) : 906 – 911.
[27] 边志强, 程卫强, 薛孝补, et al. 基于陀螺和星敏感器的卫星姿态确定算法[J]. 航天器工程, 2011, 20(02) : 29 – 34.
[28] LEFFERTS E J, MARKLEY F L, SHUSTER M. Unscented Filtering for Spacecraft Attitude Estimation[J]. J.guid.control Dynam, 2003, 26(4) : 536 – 542.
[29] ZHANG M D, DAI H F, HU B Q, et al. Robust adaptive UKF based on SVR for inertial based integrated navigation[J]. Defence Technology, 2019.
[30] 袁晓波, 张超, 詹银虎. 强跟踪自适应SRCKF的卫星姿态确定算法[J]. 测绘科学, 2018.
[31] 胡玉梅, ZHENTAO H, YONG J. Marginalized cubature Kalman filtering algorithm based on linear/nonlinear mixed-Gaussian model[J]. High Technology Letters, 2018, 24(4) : 28 – 34.
[32] 唐李军. Cubature卡尔曼滤波及其在导航中的应用研究[D]. [S.l.] : 哈尔滨工程大学, 2012.
[33] RAUCH H E, STRIEBEL C T, TUNG F. Maximum likelihood estimates of linear dynamic systems[J]. Aiaa Student Journal American Institute of Aeronautics & Astronautics, 1965, 3 i8(8) : 1445 – 1450.
[34] 鲍雨波, 宗红, 张春青. RTS平滑滤波在事后姿态确定中的应用[J]. 空间控制技术与应用, 2015, 41(3) : 18 – 18.
[35] WU X T, YANG Y, GAO W, et al. Attitude error compensation technology for inertial platform of airborne gravimeter based on RTS smoothing algorithm[C] 2017 24th Saint Petersburg International Conference on Integrated Navigation Systems (ICINS). 2017.
[36] 刘一武, 陈义庆. 星敏感器测量模型及其在卫星姿态确定系统中的应用[J]. 宇航学报, 2003(02) : 49 – 54.
[37] 王淑一, 刘祥, 陈超. 资源三号卫星控制系统概述及在轨验证[J]. 空间控制技术与应用, 2015, 41(2).
[38] KIEFFER P, ARMBRUSTER P, GERNER J L, et al. European Cooperation for Space Standardization (ECSS) Space Communication Standards[C] // Dasia-data Systems in Aerospace. 2004.
[39] BLARRE L, OUAKNINE J, ODDOS-MARCEL L, et al. High Accuracy Sodern Star Trackers: Recent Improvements Proposed on SED36 and HYDRA Star Trackers[C] // AIAA Guidance, Navigation, and Control Conference and Exhibit. 2006.
[40] JORGENSEN P S, JORGENSEN J L, DENVER T, et al. In-flight quality and accuracy of attitude measurements from the CHAMP advanced stellar compass[J]. Acta Astronautica, 2005, 56(1-2) : 181 – 186.
[41] IGARASHI T. Alos mission requirement and sensor specifications[J]. Advances in Space Research, 2001, 28(1) : 127 – 131.
[42] 王兴涛, 李迎春, 李晓燕. ”天绘一号”卫星星敏感器精度分析[J]. 遥感学报, 2012, 16(0z1) : 90 – 93.
[43] 庞博, 黎康, 汤亮, et al. 卫星星敏感器精度分析[J]. 空间控制技术与应用, 2017, 43(01) : 17 – 24.
[44] 卢欣, 武延鹏, 钟红军, et al. 星敏感器低频误差分析[J]. 空间控制技术与应用, 2014, 40(02) : 1 – 7.
[45] 牛睿, 朱文山, 李利亮, et al. 高分五号卫星姿轨控分系统设计[J]. 上海航天, 2019(S2).
[46] LAI Y, GU D, LIU J, et al. Low-Frequency Error Extraction and Compensation for Attitude Measurements from STECE Star Tracker[J]. Sensors, 2016, 16(10) : 1669.
[47] AIAA. Filtering Methods for Error Reduction in Spacecraft Attitude Estimation Using Quaternion Star Trackers - AIAA Guidance, Navigation, and Control Conference (AIAA)[J]. Aiaa Journal, 2011.
[48] XIONG K, ZHANG C Q, LIU L D. Identification of star sensor low-frequency error parameters[J]. IET Control Theory and Applications, 2012, 6(3) : 384 – 393.
[49] 雷琦, 龚志辉, 林雕. AEKF在星敏感器低频误差补偿中的应用[J]. 测绘科学技术学报, 2016, 038(003) : 252 – 257.
[50] 熊凯, 汤亮, 刘一武. 基于地标信息的星敏感器低频误差标定方法[J]. 空间控制技术与应用, 2012, 038(003) : 11 – 15.
[51] 庞博, 李果, 黎康, et al. 一种基于地标的星敏感器低频误差在轨校正方法[J]. 航天器工程, 2018, v.27;No.130(03) : 83 – 89.
[52] 王任享, 王建荣, 胡莘. 在轨卫星无地面控制点摄影测量探讨[J]. 武汉大学学报:信息科学版, (11) : 1261 – 1264.
[53] 王任享, 王建荣, 胡莘. 卫星摄影姿态测定系统低频误差补偿[J]. 测绘学报, 2016, 45(02) : 127 – 130.
[54] 吕振铎, 雷拥军. 卫星姿态测量与确定[M]. [S.l.] : 卫星姿态测量与确定, 2013.
[55] SOLà J. Quaternion kinematics for the error-state Kalman filter[J], 2017.
[56] ANON. State estimation for robotics[J], 2017.