载人飞船自主应急返回动力学与控制研究
|
摘要
随着人类迈入外空间进程的快速发展,载人航天中的安全问题日益突出。过去在“SZ-4号”至“SZ-7号”飞船上试验的自主应急返回控制方案是在地面精密定轨和地面可测控条件下进行返回控制参数注入,从而实现在轨多圈的应急返回控制的。
载人航天交会对接二期工程正在实施过程中,一旦在飞船交会对接变轨的过程中出现紧急情况,又没有地面测控支承条件,以前“SZ-4号”至“SZ-7号”的自主应急返回控制方案和算法不再有效。本文正是在此工程背景下完成了飞船自主应急返回船上方案和算法的设计,以及与地面数值仿真的精度比对,结果表明满足工程技术设计要求,可以完全不需要地面测控支持,实现了完全的自主应急返回控制。飞船自主应急返回要求自主定轨、对返回控制参数和落点预报的计算全部由船载计算机实时计算完成。受船载计算机硬件和实时计算的限制,飞船自主应急返回方案的算法必须有很高的计算效率。
本文对飞船自主应急返回的研究内容包括飞船在轨运行段的自主定轨、轨道预报动力学、制动返回段的神经网络控制算法以及提供给飞船的地面测试仿真软件四部分。飞船自主定轨基于GPS数据完成,在轨运行段的轨道预报采用考虑大气阻力摄动修正的拟平根数解析算法,并通过已有的GPS数据对归一化大气阻力修正系数进行辨识,仿真结果表明轨道预报算法有较高的计算效率和预报精度。飞船制动返回段采用基于BP人工神经网络的算法,根据地面进行飞船返回再入段六自由度仿真得到的大量返回控制参数样本,然后用这些样本对建立好的BP网络进行训练;在飞船发射前将训练好的网络拟合参数装订到飞船上,从而实现飞船返回控制参数与落点的实时计算。
通过飞船自主应急返回地面仿真软件的大量仿真算例表明,本文所研究的飞船自主应急返回方案是合理可行的。通过GPS实测数据与本文轨道预报算法的比较,表明本文的轨道预报算法有较高的精度,可以满足飞船自主返回控制的要求。最后,针对目前飞船返回再入BP神经拟合网络算法难以适应轨道倾角大幅度变化的局限性和不足,又初步提出了一种基于飞船轨道根数的返回再入落点半解析预估概念和算法,可较好克服神经网络算法的不足,也是今后的一个研究方向。本文提出的飞船返回再入落点半解析预估算法包括制动段、自由飞行段忽略大气阻力的解析算法和再入段插值的半解析算法。
With the quick development of human race’s process of stepping into the outer space, the security issues in the manned space flight become increasingly prominent. In the past, the control precept of autonomous emergency return tested on‘SZ-4’to‘SZ-7’is based on the injection of control parameters of return with the ground supports of measurement ,control, and precision orbit determination based on numerical calculation.
The second phase of manned space rendezvous is on scheme. Once any emergency happened in the process of rendezvous and dock, there may be no ground-based support and monitoring, the control scheme and algorithm of autonomous emergency return performed on‘SZ-4’to‘SZ-7’is no longer valid. In the context of this project, this paper completed the design of control scheme and algorithm of the spaceship’s autonomous emergency retun. By comparing with the numerical simulation, it has been proved that this control scheme and algorithm can meet the engineering design requirements, and can realize autonomous emergency return reliably without the ground supports. The spaceship’s autonomous emergency return requires that the autonomous orbit determination, the prediction of control parameters of return and the prediction of the fall point must be calculated in real-time by the computer onboard. By constraint of the hardware of the computer onboard, the algorithm of autonomous ermergency return must have a very high computational efficiency.
The dynamics and control research of spacecraft’s autonomous emergency return studied in this paper have been divided into four parts, the autonomous orbit determination, orbit prediction, neural network algorithm of returning control parameters and ground-tested simulation software for spacecraft. The orbit prediction used the algothrim of mean element considering the perturbation due to atmospheric drag, and the normalized coefficient indentification of atmospheric drag was based on the existed GPS datas. The simulation results have showed that the algothrim of orbit prediction has been high computational efficiency and forcast accuracy. The BP artifical neural network algorithm has been adopted in the braking back segment. By the 6-DOF simulations on ground of spacecraft’s return re-entry, a large number of samples of flight parameters had been gathered for the training of the BP neural network. Before the launch of the spacecraft, the parameters of the network which had been trained well would be fixed in it, allowing the spacecraft to calculate the return control parameters and the points of fall in real time.
According to a large number of simulative examples finished by the software of the spacecraft’s autonomous emergency return simulation, the precept of spacecraft’s autonomous emergency return studied in this paper is reasonable and feasible. By comparing the real GPS datas with the results of the algorithm of orbit prediction, it has showed that the algorithm of orbit prediction with high accuracy, can meet the requirements of spacecraft’s autonomous return control.
In the end, in view of the limitations and disadvantages of current algorithm of spacecraft’s re-entry BP neural network that it was difficult to adapt to the large changes of orbital inclination, a new semi-analytical concept and algorithm about the prediction of the points of fall based on orbital elements has been put forward. The new algorithm may overcome the disadvantages of the algorithm of BP neural network, and it will also be a future research direction. This new kind of semi-analytical algorithm about the fall point prediction includes the analytic algorithm with neglected atmospheric drag in braking section and free-flying section, and the semi-analytical interpolation algorithm in re-entry section.
载人航天交会对接二期工程正在实施过程中,一旦在飞船交会对接变轨的过程中出现紧急情况,又没有地面测控支承条件,以前“SZ-4号”至“SZ-7号”的自主应急返回控制方案和算法不再有效。本文正是在此工程背景下完成了飞船自主应急返回船上方案和算法的设计,以及与地面数值仿真的精度比对,结果表明满足工程技术设计要求,可以完全不需要地面测控支持,实现了完全的自主应急返回控制。飞船自主应急返回要求自主定轨、对返回控制参数和落点预报的计算全部由船载计算机实时计算完成。受船载计算机硬件和实时计算的限制,飞船自主应急返回方案的算法必须有很高的计算效率。
本文对飞船自主应急返回的研究内容包括飞船在轨运行段的自主定轨、轨道预报动力学、制动返回段的神经网络控制算法以及提供给飞船的地面测试仿真软件四部分。飞船自主定轨基于GPS数据完成,在轨运行段的轨道预报采用考虑大气阻力摄动修正的拟平根数解析算法,并通过已有的GPS数据对归一化大气阻力修正系数进行辨识,仿真结果表明轨道预报算法有较高的计算效率和预报精度。飞船制动返回段采用基于BP人工神经网络的算法,根据地面进行飞船返回再入段六自由度仿真得到的大量返回控制参数样本,然后用这些样本对建立好的BP网络进行训练;在飞船发射前将训练好的网络拟合参数装订到飞船上,从而实现飞船返回控制参数与落点的实时计算。
通过飞船自主应急返回地面仿真软件的大量仿真算例表明,本文所研究的飞船自主应急返回方案是合理可行的。通过GPS实测数据与本文轨道预报算法的比较,表明本文的轨道预报算法有较高的精度,可以满足飞船自主返回控制的要求。最后,针对目前飞船返回再入BP神经拟合网络算法难以适应轨道倾角大幅度变化的局限性和不足,又初步提出了一种基于飞船轨道根数的返回再入落点半解析预估概念和算法,可较好克服神经网络算法的不足,也是今后的一个研究方向。本文提出的飞船返回再入落点半解析预估算法包括制动段、自由飞行段忽略大气阻力的解析算法和再入段插值的半解析算法。
With the quick development of human race’s process of stepping into the outer space, the security issues in the manned space flight become increasingly prominent. In the past, the control precept of autonomous emergency return tested on‘SZ-4’to‘SZ-7’is based on the injection of control parameters of return with the ground supports of measurement ,control, and precision orbit determination based on numerical calculation.
The second phase of manned space rendezvous is on scheme. Once any emergency happened in the process of rendezvous and dock, there may be no ground-based support and monitoring, the control scheme and algorithm of autonomous emergency return performed on‘SZ-4’to‘SZ-7’is no longer valid. In the context of this project, this paper completed the design of control scheme and algorithm of the spaceship’s autonomous emergency retun. By comparing with the numerical simulation, it has been proved that this control scheme and algorithm can meet the engineering design requirements, and can realize autonomous emergency return reliably without the ground supports. The spaceship’s autonomous emergency return requires that the autonomous orbit determination, the prediction of control parameters of return and the prediction of the fall point must be calculated in real-time by the computer onboard. By constraint of the hardware of the computer onboard, the algorithm of autonomous ermergency return must have a very high computational efficiency.
The dynamics and control research of spacecraft’s autonomous emergency return studied in this paper have been divided into four parts, the autonomous orbit determination, orbit prediction, neural network algorithm of returning control parameters and ground-tested simulation software for spacecraft. The orbit prediction used the algothrim of mean element considering the perturbation due to atmospheric drag, and the normalized coefficient indentification of atmospheric drag was based on the existed GPS datas. The simulation results have showed that the algothrim of orbit prediction has been high computational efficiency and forcast accuracy. The BP artifical neural network algorithm has been adopted in the braking back segment. By the 6-DOF simulations on ground of spacecraft’s return re-entry, a large number of samples of flight parameters had been gathered for the training of the BP neural network. Before the launch of the spacecraft, the parameters of the network which had been trained well would be fixed in it, allowing the spacecraft to calculate the return control parameters and the points of fall in real time.
According to a large number of simulative examples finished by the software of the spacecraft’s autonomous emergency return simulation, the precept of spacecraft’s autonomous emergency return studied in this paper is reasonable and feasible. By comparing the real GPS datas with the results of the algorithm of orbit prediction, it has showed that the algorithm of orbit prediction with high accuracy, can meet the requirements of spacecraft’s autonomous return control.
In the end, in view of the limitations and disadvantages of current algorithm of spacecraft’s re-entry BP neural network that it was difficult to adapt to the large changes of orbital inclination, a new semi-analytical concept and algorithm about the prediction of the points of fall based on orbital elements has been put forward. The new algorithm may overcome the disadvantages of the algorithm of BP neural network, and it will also be a future research direction. This new kind of semi-analytical algorithm about the fall point prediction includes the analytic algorithm with neglected atmospheric drag in braking section and free-flying section, and the semi-analytical interpolation algorithm in re-entry section.
引文
[1]张育林,郑荣跃,沈力平,陈善广.载人航天工程基础.长沙:国防科技大学出版社,1997
[2]李思强.载人飞船工程学概述.北京:科学出版社,1985
[3]林聪榕等.国外载人航天技术的发展.国防科技参考,14(3)
[4]蔡道济等.美国和前苏联载人航天飞船返回和回收研究.国防科技信息中心报告(94-04-01-02),1994
[5]张健等.联盟号飞船总体设计.航天器工程,1994,3(2)
[6]孙悦年.联盟号飞船的典型机构设计.航天器工程,1994,3(2)
[7]王丹阳.载人飞船联盟号的十一种状态.中国航天,1993(3)
[8]王金华.“联盟-TM”飞船上的生命保障系统概述.载人航天,1995(1)
[9]田莉等编译.前苏联载人航天活动.航天总公司第五研究院,1995
[10]陈国良,高树义.中国航天器回收着陆技术50年成就与展望.航天返回与遥感,2008(9)
[11]李颐黎,张书庭,柯伦.神舟飞船应急救生分系统研制与飞行结果评价.航天器工程,2004(3)
[12]童旭东,李惠康,葛玉君,高树义.神舟号载人飞船回收着陆系统设计与性能评估.航天器工程,2004(3)
[13]唐伯昶,陈国良.中国返回式卫星发展与成就.航天返回与遥感,2008(9)
[14]侯黎强,谭炜,路平,韩明.飞船自主应急返回控制参数确定方法及应用.飞行器测控学报,2006(25)
[15]赵汉元,载人飞船再入轨道和制导规律设计及六自由度弹道仿真.航天返回与遥感,2001(3)
[16]汤国建,赵汉元.载人飞船返回舱六自由度再入弹道仿真研究.飞行力学,2000(3)
[17]白东升.载人飞船应急返回轨迹规划设计.硕士学位论文,北京航空航天大学,2000
[18]吴启星.载人飞船应急救生轨道的返回再入动力学与制导问题研究.硕士学位论文,国防科学技术大学,2002
[19]胡正东,李鹏奎,张士峰,蔡洪.载人航天器高空应急救生控制变量与落点的快速预报算法.宇航学报,2007(11)
[20]郭叔伟,董杨彪,丁娣,秦子增.载人航天器低空救生仿真.国防科技大学学报,2008(4)
[21]汤锡生,陈贻迎,朱民才.载人飞船轨道确定和返回控制.北京:国防工业出版社,2002
[22]郗晓宁,王威,高玉东.近地航天器轨道基础.长沙:国防科学技术大学,2003
[23]刘林.航天器轨道理论.北京:国防工业出版社,2000
[24]紫金山天文台编.中国天文年历.北京:科学出版社,1995
[25]范剑峰.空间站工程概论.哈尔滨:哈尔滨工业大学出版社,1990
[26]韩洪硕.飞船系统的发展研究.航天部707所,1992
[27] Garfinkel B. Formal solution in the problem of small divisors. Astron. J. 1966, 71(8)
[28] Garfinkel B, Jupp A H and Williams C A. Arecursive Von Zeipel algorithm for the ideal resonance problem. Astron.J. 1971, 76(2)
[29] Kienappel K, Koppenwallner G and Legge H. Rarefied Gas Dynamics. In Proceeding of the Eighth International Symposium (1972)
[30] Taff L G.Computational Spherical Astronomy. New York: John Wiley and Sons, Inc. 1981
[31]任萱.自由飞行时摄动方程的状态转移矩阵的解析解.中国空间科学技术,1983(1)
[32]郑伟,汤国建.弹道导弹自由段解算的等高约束解析解.宇航学报,2007(2)
[33]潘刚.远程弹道导弹快速诸元计算方法研究.硕士学位论文,国防科学技术大学,2002
[34]朱龙根.改进的Barrar型中间轨道.国防科学技术大学学报,1985(2)
[35]李振涛.自由飞行轨道解析解的中间轨道方法.导弹与航天运载技术,1993(5)
[36]李晓明.经典f、g级数的修正法.国防科技大学学报,1991(2)
[37]张毅.弹道导弹弹头再入参数的解析解.国际弹道学会议(ICB).南京.1988
[38]李连仲.弹道飞行器自由飞行轨道的解析解法.宇航学报,1982(1)
[39] King-Hele D G and Walker D M C. Vpper-atmosphere zonal winds from satellite orbit analysis. Planet. Space Sci. 1983, 31(5)
[40] Brouwer D and Hori G. Theoretical evalution of atmospheric drag effects in the motion of an artifical satellite. Astron.J. 1961, 66(4)
[41] Tapley B D. Statistical orbit determination theory. In Proceeding of NATO Adavanced Study Institute in Dynamical Astronomy. Italy, Reided publishing Co.Dordrecht, 1976
[42] Green R M.Spherical Astronomy. New York: Combridge University Press, 1985
[43]冯书兴,张守信.载人航天器返回舱落点预报及精度分析.长沙:国防科技大学出版社,1996
[44]赵汉元.飞行器再入动力学和制导.长沙:国防科技大学出版社,1997
[45]何力,赵汉元.载人飞船标准返回轨道设计.国防科技大学学报,1996(3)
[46]赵汉元,陈克俊,郭振云,程国采.再入机动飞行器数学仿真研究.宇航学报,1997(1)
[47]贾沛然,陈克俊,何力.远程火箭弹道学.长沙:国防科技大学出版社,1993
[48]王希季,李颐黎.航天器进入与返回技术.北京:宇航出版社,1991
[49]程国采.四元数法及其应用.长沙:国防科技大学出版社,1991
[50]刘林,王彦荣.卫星轨道预报的一种分析方法.天文学报,2005(7)
[51]刘林,廖新浩.关于数值求解人造卫星运动方程的几个问题.天文学报,1997,38(1)
[52]刘林,Innanen K A.关于人造卫星运动中的临界角和通约问题.天文学报,1986,27(1)
[53]夏灿英,吴玉媛,刘林.考虑阻力系数变化的近地卫星轨道的联合摄动.天文学报,1982,23(2)
[54]刘君,夏智勋.动力学系统辨识与建模.长沙:国防科技大学出版社,2007
[55]张如伟,刘根有.低轨卫星轨道拟合及预报方法研究.大地测量与地球动力学,2008(8)
[56] Guikirpal S, Fred Y H.Autonomous path-planning for formation-flying applications. Jet Propulsion Laboratory. 2002
[57] Thompson D C, Raitt W J, Bonifazi C, Williams S D, Aguero V M, Gilchrist B E, Banks P M. TSS-1-Orbiter current and voltage experiments. 31st Aerospace Science Meeting & Exhibit, January 11-14, 1993, Reno, Nevada. AIAA-93-0702
[58] Yoshida Kazuya, Hashizume Kenihi, Nenhev Dragomir N, Inaba Noriyasu, Oda Mitsushige. Control of a space manipulator autonomous target apture-ETS-VII flight experiments and analysis. AIAA Guidance, Navigation, and Control Conference and Exhibit, Denver, CO, August 14-17, 2000. AIAA-2000-4376
[59] Uchiyama Kenji, Fujii Hironori A, Yoneoka Hideharu. Experiments on dynamic behavior of space manipulator using suspension system. AIAA Guidance, Navigation and Control Conference, Baltimore, MD, August 7-10, 1995, AIAA-95-3334
[60]陈祥光,裴旭东.人工神经网络技术及应用.北京:中国电力出版社,2003
[61]飞思科技产品研发中心.神经网络理论与MATLAB 7实现.北京:电子工业出版社,2005
[62]崔富章.基于BP神经网络的空间飞行器姿态控制研究.硕士学位论文,重庆大学,2007
[63] Special issue on neural networks for data mining and knowledge discovery. IEEE Trans. on Neural Networks, May 2000, 11(3)
[64] Haykin Simon. Neural networks expand horizons. IEEE signal Processing Magazine, 1996, 6(2)
[65] Oja E. A simplified neuron model as a principle component analyzer. Journal of Mathmatical Biology, 1982, 15
[66] Hartman E J, et al. Layered neural networks with Gaussian hidden units as universal approximation. Neural Computation, 1990, 2(2)
[67] Park I, et al. Approxmation and radial basis function networks. Neural Computation, 1993, 5
[68] Park Y R, et al. Prediction sun spots using layered perceptron neural networks. IEEE trans. On Neural Networks, 1996, 7(2)
[69] McCulloch W S, Pitts W. A logical calculus of the ideas immanent in neurous activity. Bulletin of Mathmatical Biophysics, 1943, 6
[70]谭浩强编著.C++程序设计.北京:清华大学出版社,2004
[71]宋旭民,秦子增,程文科,彭勇.航天器回收着陆仿真软件系统(ARLSSS)简介.航天返回与遥感,2004(9)
[72]刘林.人造卫星轨道力学.北京:高等教育出版社,1992
[73]刘新建等.XX-X自主应急返回轨道预报和返回控制的数值仿真研究.国防科技大学航天与材料工程学院,2009
[2]李思强.载人飞船工程学概述.北京:科学出版社,1985
[3]林聪榕等.国外载人航天技术的发展.国防科技参考,14(3)
[4]蔡道济等.美国和前苏联载人航天飞船返回和回收研究.国防科技信息中心报告(94-04-01-02),1994
[5]张健等.联盟号飞船总体设计.航天器工程,1994,3(2)
[6]孙悦年.联盟号飞船的典型机构设计.航天器工程,1994,3(2)
[7]王丹阳.载人飞船联盟号的十一种状态.中国航天,1993(3)
[8]王金华.“联盟-TM”飞船上的生命保障系统概述.载人航天,1995(1)
[9]田莉等编译.前苏联载人航天活动.航天总公司第五研究院,1995
[10]陈国良,高树义.中国航天器回收着陆技术50年成就与展望.航天返回与遥感,2008(9)
[11]李颐黎,张书庭,柯伦.神舟飞船应急救生分系统研制与飞行结果评价.航天器工程,2004(3)
[12]童旭东,李惠康,葛玉君,高树义.神舟号载人飞船回收着陆系统设计与性能评估.航天器工程,2004(3)
[13]唐伯昶,陈国良.中国返回式卫星发展与成就.航天返回与遥感,2008(9)
[14]侯黎强,谭炜,路平,韩明.飞船自主应急返回控制参数确定方法及应用.飞行器测控学报,2006(25)
[15]赵汉元,载人飞船再入轨道和制导规律设计及六自由度弹道仿真.航天返回与遥感,2001(3)
[16]汤国建,赵汉元.载人飞船返回舱六自由度再入弹道仿真研究.飞行力学,2000(3)
[17]白东升.载人飞船应急返回轨迹规划设计.硕士学位论文,北京航空航天大学,2000
[18]吴启星.载人飞船应急救生轨道的返回再入动力学与制导问题研究.硕士学位论文,国防科学技术大学,2002
[19]胡正东,李鹏奎,张士峰,蔡洪.载人航天器高空应急救生控制变量与落点的快速预报算法.宇航学报,2007(11)
[20]郭叔伟,董杨彪,丁娣,秦子增.载人航天器低空救生仿真.国防科技大学学报,2008(4)
[21]汤锡生,陈贻迎,朱民才.载人飞船轨道确定和返回控制.北京:国防工业出版社,2002
[22]郗晓宁,王威,高玉东.近地航天器轨道基础.长沙:国防科学技术大学,2003
[23]刘林.航天器轨道理论.北京:国防工业出版社,2000
[24]紫金山天文台编.中国天文年历.北京:科学出版社,1995
[25]范剑峰.空间站工程概论.哈尔滨:哈尔滨工业大学出版社,1990
[26]韩洪硕.飞船系统的发展研究.航天部707所,1992
[27] Garfinkel B. Formal solution in the problem of small divisors. Astron. J. 1966, 71(8)
[28] Garfinkel B, Jupp A H and Williams C A. Arecursive Von Zeipel algorithm for the ideal resonance problem. Astron.J. 1971, 76(2)
[29] Kienappel K, Koppenwallner G and Legge H. Rarefied Gas Dynamics. In Proceeding of the Eighth International Symposium (1972)
[30] Taff L G.Computational Spherical Astronomy. New York: John Wiley and Sons, Inc. 1981
[31]任萱.自由飞行时摄动方程的状态转移矩阵的解析解.中国空间科学技术,1983(1)
[32]郑伟,汤国建.弹道导弹自由段解算的等高约束解析解.宇航学报,2007(2)
[33]潘刚.远程弹道导弹快速诸元计算方法研究.硕士学位论文,国防科学技术大学,2002
[34]朱龙根.改进的Barrar型中间轨道.国防科学技术大学学报,1985(2)
[35]李振涛.自由飞行轨道解析解的中间轨道方法.导弹与航天运载技术,1993(5)
[36]李晓明.经典f、g级数的修正法.国防科技大学学报,1991(2)
[37]张毅.弹道导弹弹头再入参数的解析解.国际弹道学会议(ICB).南京.1988
[38]李连仲.弹道飞行器自由飞行轨道的解析解法.宇航学报,1982(1)
[39] King-Hele D G and Walker D M C. Vpper-atmosphere zonal winds from satellite orbit analysis. Planet. Space Sci. 1983, 31(5)
[40] Brouwer D and Hori G. Theoretical evalution of atmospheric drag effects in the motion of an artifical satellite. Astron.J. 1961, 66(4)
[41] Tapley B D. Statistical orbit determination theory. In Proceeding of NATO Adavanced Study Institute in Dynamical Astronomy. Italy, Reided publishing Co.Dordrecht, 1976
[42] Green R M.Spherical Astronomy. New York: Combridge University Press, 1985
[43]冯书兴,张守信.载人航天器返回舱落点预报及精度分析.长沙:国防科技大学出版社,1996
[44]赵汉元.飞行器再入动力学和制导.长沙:国防科技大学出版社,1997
[45]何力,赵汉元.载人飞船标准返回轨道设计.国防科技大学学报,1996(3)
[46]赵汉元,陈克俊,郭振云,程国采.再入机动飞行器数学仿真研究.宇航学报,1997(1)
[47]贾沛然,陈克俊,何力.远程火箭弹道学.长沙:国防科技大学出版社,1993
[48]王希季,李颐黎.航天器进入与返回技术.北京:宇航出版社,1991
[49]程国采.四元数法及其应用.长沙:国防科技大学出版社,1991
[50]刘林,王彦荣.卫星轨道预报的一种分析方法.天文学报,2005(7)
[51]刘林,廖新浩.关于数值求解人造卫星运动方程的几个问题.天文学报,1997,38(1)
[52]刘林,Innanen K A.关于人造卫星运动中的临界角和通约问题.天文学报,1986,27(1)
[53]夏灿英,吴玉媛,刘林.考虑阻力系数变化的近地卫星轨道的联合摄动.天文学报,1982,23(2)
[54]刘君,夏智勋.动力学系统辨识与建模.长沙:国防科技大学出版社,2007
[55]张如伟,刘根有.低轨卫星轨道拟合及预报方法研究.大地测量与地球动力学,2008(8)
[56] Guikirpal S, Fred Y H.Autonomous path-planning for formation-flying applications. Jet Propulsion Laboratory. 2002
[57] Thompson D C, Raitt W J, Bonifazi C, Williams S D, Aguero V M, Gilchrist B E, Banks P M. TSS-1-Orbiter current and voltage experiments. 31st Aerospace Science Meeting & Exhibit, January 11-14, 1993, Reno, Nevada. AIAA-93-0702
[58] Yoshida Kazuya, Hashizume Kenihi, Nenhev Dragomir N, Inaba Noriyasu, Oda Mitsushige. Control of a space manipulator autonomous target apture-ETS-VII flight experiments and analysis. AIAA Guidance, Navigation, and Control Conference and Exhibit, Denver, CO, August 14-17, 2000. AIAA-2000-4376
[59] Uchiyama Kenji, Fujii Hironori A, Yoneoka Hideharu. Experiments on dynamic behavior of space manipulator using suspension system. AIAA Guidance, Navigation and Control Conference, Baltimore, MD, August 7-10, 1995, AIAA-95-3334
[60]陈祥光,裴旭东.人工神经网络技术及应用.北京:中国电力出版社,2003
[61]飞思科技产品研发中心.神经网络理论与MATLAB 7实现.北京:电子工业出版社,2005
[62]崔富章.基于BP神经网络的空间飞行器姿态控制研究.硕士学位论文,重庆大学,2007
[63] Special issue on neural networks for data mining and knowledge discovery. IEEE Trans. on Neural Networks, May 2000, 11(3)
[64] Haykin Simon. Neural networks expand horizons. IEEE signal Processing Magazine, 1996, 6(2)
[65] Oja E. A simplified neuron model as a principle component analyzer. Journal of Mathmatical Biology, 1982, 15
[66] Hartman E J, et al. Layered neural networks with Gaussian hidden units as universal approximation. Neural Computation, 1990, 2(2)
[67] Park I, et al. Approxmation and radial basis function networks. Neural Computation, 1993, 5
[68] Park Y R, et al. Prediction sun spots using layered perceptron neural networks. IEEE trans. On Neural Networks, 1996, 7(2)
[69] McCulloch W S, Pitts W. A logical calculus of the ideas immanent in neurous activity. Bulletin of Mathmatical Biophysics, 1943, 6
[70]谭浩强编著.C++程序设计.北京:清华大学出版社,2004
[71]宋旭民,秦子增,程文科,彭勇.航天器回收着陆仿真软件系统(ARLSSS)简介.航天返回与遥感,2004(9)
[72]刘林.人造卫星轨道力学.北京:高等教育出版社,1992
[73]刘新建等.XX-X自主应急返回轨道预报和返回控制的数值仿真研究.国防科技大学航天与材料工程学院,2009