基于神经网络的无人机动态逆自主飞行控制系统研究
|
摘要
无人机在现代战场中发挥着越来越大的作用。由于其具有机动性高、零伤亡和低成本等独特的优点,世界各国对无人机的发展给予了极大的关注。随着无人机在军民用领域应用的日益广泛,人们对无人机飞行性能的要求越来越高,赋予无人机智能以实现其自主飞行的呼声越来越大。现代计算机信息技术、人工智能和自主控制技术的发展,指出了无人机自主飞行控制发展的方向。
本文研究的主要内容是根据无人机自主飞行的控制要求,设计基于神经网络补偿的无人机动态逆自主飞行控制系统。首先介绍了无人机的发展历史和国内外发展现状,分析了未来无人机面向自主控制的发展趋势。其次总结了传统控制方法在无人机自主飞行控制研究中的不足,介绍了现代先进的飞行控制律设计方法。然后运用人工神经网络和逆系统控制方法,从内外回路两个方面,提出了样例无人直升机自主飞行控制系统基本结构。基于该结构,根据动态逆反馈线性化理论设计无人机内回路姿态控制系统和外回路轨迹控制系统,并分析模型跟踪误差动力特性,再根据误差特性设计神经网络误差补偿器,给出神经网络在线学习调整算法。最后运用VS程序设计工具开发了无人机自主飞行控制仿真软件,对控制系统理论的正确性和合理性进行仿真实验验证,并参照飞行品质规范进行了典型任务科目的飞行仿真实验。仿真结果表明本文设计的控制系统具有较高的控制精度和自主适应性,符合自主飞行控制要求。
Nowadays, the UAV plays a more and more important role in modern wars. Because of its high maneuverability, zero death and low cost, countries all around the world put much effort into studying it. As the application of UAV in the civil fields is becoming more and more broad, people pays more attention to its flight performance. So it is urgent to achieve the autonomous flight. The development of the computer and information technology, artificial intelligence and autonomous control technology promote the further studying on the UAV’s autonomous flight control system.
The main content of this paper is to design the UAV’s autonomous flight control system, according to the UAV’s control requirements. First, the history and development of the UAV is introduced, and the development tendency of the UAV’s autonomous flight control system is analyzed as well. Then, the traditional control methods are described. Based on their disadvantages in the studying of the UAV’s autonomous flight control system, the advanced control methods are proposed, such as the dynamical inversion flight control method. Using this method, the basic structure of the UAV’s autonomous flight control system is conducted, from the respects of the inner-loop and the outer-loop. According to this structure, the inner-loop attitude control system and the trajectory control system are designed respectively. Error compensation system based on Neural Network is developed according to the characteristics of model tracking error. Besides, online learning algorithm of the Neural Network is given. Finally, implementing the VS programming language, the UAV’s autonomous flight control and simulation software is developed. Within the simulation environment, the correctness and rationality is validated, and the simulation experiment of typical flight task is performed reference to the flight quality standard. The simulation result shows that the control system designed in the paper has high control accuracy as well as adaptability,and can meet the requirements of the autonomous flight well.
本文研究的主要内容是根据无人机自主飞行的控制要求,设计基于神经网络补偿的无人机动态逆自主飞行控制系统。首先介绍了无人机的发展历史和国内外发展现状,分析了未来无人机面向自主控制的发展趋势。其次总结了传统控制方法在无人机自主飞行控制研究中的不足,介绍了现代先进的飞行控制律设计方法。然后运用人工神经网络和逆系统控制方法,从内外回路两个方面,提出了样例无人直升机自主飞行控制系统基本结构。基于该结构,根据动态逆反馈线性化理论设计无人机内回路姿态控制系统和外回路轨迹控制系统,并分析模型跟踪误差动力特性,再根据误差特性设计神经网络误差补偿器,给出神经网络在线学习调整算法。最后运用VS程序设计工具开发了无人机自主飞行控制仿真软件,对控制系统理论的正确性和合理性进行仿真实验验证,并参照飞行品质规范进行了典型任务科目的飞行仿真实验。仿真结果表明本文设计的控制系统具有较高的控制精度和自主适应性,符合自主飞行控制要求。
Nowadays, the UAV plays a more and more important role in modern wars. Because of its high maneuverability, zero death and low cost, countries all around the world put much effort into studying it. As the application of UAV in the civil fields is becoming more and more broad, people pays more attention to its flight performance. So it is urgent to achieve the autonomous flight. The development of the computer and information technology, artificial intelligence and autonomous control technology promote the further studying on the UAV’s autonomous flight control system.
The main content of this paper is to design the UAV’s autonomous flight control system, according to the UAV’s control requirements. First, the history and development of the UAV is introduced, and the development tendency of the UAV’s autonomous flight control system is analyzed as well. Then, the traditional control methods are described. Based on their disadvantages in the studying of the UAV’s autonomous flight control system, the advanced control methods are proposed, such as the dynamical inversion flight control method. Using this method, the basic structure of the UAV’s autonomous flight control system is conducted, from the respects of the inner-loop and the outer-loop. According to this structure, the inner-loop attitude control system and the trajectory control system are designed respectively. Error compensation system based on Neural Network is developed according to the characteristics of model tracking error. Besides, online learning algorithm of the Neural Network is given. Finally, implementing the VS programming language, the UAV’s autonomous flight control and simulation software is developed. Within the simulation environment, the correctness and rationality is validated, and the simulation experiment of typical flight task is performed reference to the flight quality standard. The simulation result shows that the control system designed in the paper has high control accuracy as well as adaptability,and can meet the requirements of the autonomous flight well.
引文
[1] Office of the Secretary of Defense.Unmanned Aircraft Systems Roadmap,2005-2030.Washington DC,2005.
[2]陈海.无人机自主控制综述及自主着陆控制系统设计[M].西北工业大学硕士论文, 2007.
[3] http://baike.baidu.com/view/120502.htm
[4]刘波.无人机非线性姿态控制律设计及仿真研究[M].中南大学硕士论文, 2006.
[5] http://mil.news.sina.com.cn/2004-03-25/0836189310.html
[6] http://baike.baidu.com/view/19743.htm
[7]王辉.无人直升机自主飞行控制系统设计与实现[D].南京航空航天大学博士论文, 2005.
[8] Napolitano M.R., Kincheloe M. On-line Learning Neural-Network Controllers for Autopilot System [J]. Journal of Guidance Control and Dynamics, 1995, 18(6):1008-1015.
[9] Calise A.J., Johnson E.N., Johnson M.D., etal. Applications of Adaptive Neural-Network Control to Unmanned Aerial Vehicles[R]. AIAA/ICAS International Air and Space Symposium and Exposition: The Next 100 Years, Dayton, OH, July 2003.
[10]王源.不确定非线性系统的神经网络自适应重构控制[D].南京航空航天大学博士论文, 2002.
[11]蔡燕霞.基于神经网络的专家系统在卫星故障诊断中的应用研究[D].哈尔滨工业大学硕士论文, 2001.
[12]扈宏杰.神经网络控制理论的研究及其在高精度仿真转台系统中的应用[D].北京航空航天大学博士论文, 2001.
[13]尹江辉,刘艇,王立新.模糊神经网络在过失速机动飞行中的应用[J].飞行力学, 2001,19(l):42-44.
[14] Zhaoquan Cheng. Dynamic Inversion and Model Predictive Control for Unmanned Aerial Vehicles [M]. University of Ottawa,2004.
[15] Yoonghyun Shin. Neural Network Based Adaptive Control for Nonlinear Dynamic Regimes [D]. Georgia Institute of Technology,2005.
[16] Leitner J. Helicopter Nonlinear Control Using Adaptive Feedback Linearization [D]. Georgia Institute of Technology, 1995.
[17] Sehumacher C. Adaptive Flight Control Using Dynamic Inversion and Neural Networks[J], AIAA Guidance,Navigation,and Control Conference and Exhibit, Portland,1999.
[18] Singh S.N.and Steinberg M. Adaptive Control of Feedback Linearizable Noulinear Systems WithApplication to Flight Control [J]. AIAA Guidance Navigation and Control conference, July 29-31,1996.
[19]蔡维黎.超机动飞机的非线性动态逆控制[D].南京航空航天大学博士论文,1993.
[20]郭树军.直升机超机动飞行的神经网络控制研究[D].南京航空航天大学博士论文,2000.
[21]张锐.武装直升机大机动高敏捷性神经网络鲁棒自适应飞行控制研究[D].南京航空航天大学博士论文, 2002.
[22]戴先中.多变量非线性系统的神经网络逆控制方法[M].科学出版社,2005.16-76,129-148.
[23]朱家强,朱纪洪,郭锁凤等.基于神经网络的鲁棒自适应逆飞行控制.控制理论与应用[J],2005.
[24]朱家强.基于神经网络和动态逆的超机动飞行控制技术研究[D].南京航空航天大学博士论文, 2004.
[25] Tama?s Keviczky, Ryan Ingvalson, He?ctor Rotstein, Oreste Riccardo Natale and Gary J. Balas. An Integrated Multi-Layer Approach to Software-Enabled Control: Mission Planning to Vehicle Control[R]. University of Minnesota,2004.
[26] Schrage D.P.,Vachtsevanos G., Software Enabled Control for Intelligent UAVs[A]. IEEE International Conference on Control Applications [C], 1999.
[27] Pachter M,Chandler P.R., Challenges of Autonomous Control. IEEE Control Systems Magazine, 1998.18(4):92一97.
[28] Warren R. Dufrene,Jr.. Approach for Autonomous Control of Unmanned Aerial Vehicle Using Intelligent Agents for Knowledge Creation. Digita lAvionies Systems Conference, 2004.
[29]唐强,朱志强,王建元.国外无人机自主飞行控制研究[J].系统工程与电子技术,2004.
[30]高正,陈仁良.直升机飞行动力学[M].科学出版社, 2003.
[31] Munzinger C. Development of a Real-Time Flight Simulator for an Experimental Model Helicopter [D]. Master's thesis, Georgia Institute of Technology, School of Aerospace Engineering, Atlanta, GA 30332, July 1997.
[32]杨超,张晓谷.直升机飞行动力学数学模型研究特点分析[J].飞行力学, 1998.
[33] Mettler B. Modeling Small-Scale Unmanned Rotorcraft for Advanced Flight Control Design [D]. Minnesota: Department of Mechanical Engineering Carnegie Mellon University, 2001.
[34]胡寿松.自动控制原理[M].科学出版社,2005.
[35]史忠科.线性系统理论[M].科学出版社,2008.
[36]王俊国.基于神经网络的建模方法与控制策略研究[D].华中科技大学博士论文,2004.
[37]曹梦龙,安世奇.控制系统计算机仿真技术[M].化学工业出版社,2009.
[38] Rysdyk R T, Calise A J, and R. T. N. Chen. Nonlinear Adaptive Control of Tiltrotor AircraftUsing Neural Networks [A]. AIAA/SAE World Aviation Congress [C], Anaheim,CA, 1997.
[39] Rysdyk R T, Calise A J. Adaptive Model Inversion Flight Control for Tiltrotor Aircraft [A]. American Helicopter Society Annual Forum [C], Washington, DC, May 1998.
[40] American Helicopter Society Annual Forum [C], Washington, DC, May 1998.Calise A J, Lee H, Kim N. High Bandwidth Adaptive Flight Control [A]. AIAA Guidance,Navigation and Control Conference [C], Denver, CO, August 2000.
[41]樊京,刘叔军,盖晓华,崔世林. Matlab控制系统应用与实例[M].清华大学出版社,2008,205-230.
[42]杨一栋.直升机飞行控制[M].国防工业出版社, 2007,26-56,94-107.
[43]杨军,吴希明,凡永华,袁博.倾转旋翼机飞行控制[M].航空工业出版社,2006,24-34,64-70.
[44] Hovakimyan N, Kim N, Calise A J, etal. Adaptive Output Feedback for High-Bandwidth Control of an Unmanned Helicopter [A]. AIAA Guidance, Navigation and Control Conference [C], 2001.
[45] Johnson E N, Calise A J, El-Shirbiny H A, etal. Feedback Linearization with Neural Network Augmentation Applied to X-33 Attitude Control [R].AIAA Paper:AIAA-2000-4157.
[46] Johnson E N, Calise A J. Neural Network Adaptive Control of Systems with Input Saturation [A].American Control Conference[C], Arlington, Virginia, June 2001.
[47] Calise A J. Neural Network-Based Adaptive Control for Nonlinear Flight Regimes [R],Technical Report 2002: 1-50.
[48] Karli Wastson,Christian Nagel著,齐立波,黄静译. C#入门经典[M].清华大学出版社,2006,151-220,331-378.
[49] Walter Savitch著,周靖译. C++面向对象程序设计:基础、数据结构与编程思想[M].清华大学出版社,2003.115-218,262-280.
[50]韩立群.人工神经网络教程[M].北京邮电大学出版社,2006,76-120.
[51]徐丽娜.神经网络控制[M].哈尔滨工业大学出版社,1999.116-145.
[52]张洪润.智能技术:系统设计与开发[M].北京航空航天大学出版社,2007.60-95,236-247.
[53] Lee S. Neural Network Based Adaptive Control and its Applications to Aerial Vehicles [D]. Atlanta: Georgia Institute of Technology, 2001.
[54] Johnson E N, Kannan S. Adaptive Flight Control for an Autonomous Unmanned Helicopter [R]. AIAA Paper: AIAA-2002-4439.
[55] Kannan S, Johnson E N. Adaptive Trajectory Based Flight Control for Autonomous Helicopters [A]. AIAA Digital Avionics Conference [C], Irvine, CA, October 2002.
[56] Yong Liu.Neural Adaptive Nonlinear Tracking Using Trajectory Linearization [D].Russ: College of Engineering and Technology Ohio University,2007.
[2]陈海.无人机自主控制综述及自主着陆控制系统设计[M].西北工业大学硕士论文, 2007.
[3] http://baike.baidu.com/view/120502.htm
[4]刘波.无人机非线性姿态控制律设计及仿真研究[M].中南大学硕士论文, 2006.
[5] http://mil.news.sina.com.cn/2004-03-25/0836189310.html
[6] http://baike.baidu.com/view/19743.htm
[7]王辉.无人直升机自主飞行控制系统设计与实现[D].南京航空航天大学博士论文, 2005.
[8] Napolitano M.R., Kincheloe M. On-line Learning Neural-Network Controllers for Autopilot System [J]. Journal of Guidance Control and Dynamics, 1995, 18(6):1008-1015.
[9] Calise A.J., Johnson E.N., Johnson M.D., etal. Applications of Adaptive Neural-Network Control to Unmanned Aerial Vehicles[R]. AIAA/ICAS International Air and Space Symposium and Exposition: The Next 100 Years, Dayton, OH, July 2003.
[10]王源.不确定非线性系统的神经网络自适应重构控制[D].南京航空航天大学博士论文, 2002.
[11]蔡燕霞.基于神经网络的专家系统在卫星故障诊断中的应用研究[D].哈尔滨工业大学硕士论文, 2001.
[12]扈宏杰.神经网络控制理论的研究及其在高精度仿真转台系统中的应用[D].北京航空航天大学博士论文, 2001.
[13]尹江辉,刘艇,王立新.模糊神经网络在过失速机动飞行中的应用[J].飞行力学, 2001,19(l):42-44.
[14] Zhaoquan Cheng. Dynamic Inversion and Model Predictive Control for Unmanned Aerial Vehicles [M]. University of Ottawa,2004.
[15] Yoonghyun Shin. Neural Network Based Adaptive Control for Nonlinear Dynamic Regimes [D]. Georgia Institute of Technology,2005.
[16] Leitner J. Helicopter Nonlinear Control Using Adaptive Feedback Linearization [D]. Georgia Institute of Technology, 1995.
[17] Sehumacher C. Adaptive Flight Control Using Dynamic Inversion and Neural Networks[J], AIAA Guidance,Navigation,and Control Conference and Exhibit, Portland,1999.
[18] Singh S.N.and Steinberg M. Adaptive Control of Feedback Linearizable Noulinear Systems WithApplication to Flight Control [J]. AIAA Guidance Navigation and Control conference, July 29-31,1996.
[19]蔡维黎.超机动飞机的非线性动态逆控制[D].南京航空航天大学博士论文,1993.
[20]郭树军.直升机超机动飞行的神经网络控制研究[D].南京航空航天大学博士论文,2000.
[21]张锐.武装直升机大机动高敏捷性神经网络鲁棒自适应飞行控制研究[D].南京航空航天大学博士论文, 2002.
[22]戴先中.多变量非线性系统的神经网络逆控制方法[M].科学出版社,2005.16-76,129-148.
[23]朱家强,朱纪洪,郭锁凤等.基于神经网络的鲁棒自适应逆飞行控制.控制理论与应用[J],2005.
[24]朱家强.基于神经网络和动态逆的超机动飞行控制技术研究[D].南京航空航天大学博士论文, 2004.
[25] Tama?s Keviczky, Ryan Ingvalson, He?ctor Rotstein, Oreste Riccardo Natale and Gary J. Balas. An Integrated Multi-Layer Approach to Software-Enabled Control: Mission Planning to Vehicle Control[R]. University of Minnesota,2004.
[26] Schrage D.P.,Vachtsevanos G., Software Enabled Control for Intelligent UAVs[A]. IEEE International Conference on Control Applications [C], 1999.
[27] Pachter M,Chandler P.R., Challenges of Autonomous Control. IEEE Control Systems Magazine, 1998.18(4):92一97.
[28] Warren R. Dufrene,Jr.. Approach for Autonomous Control of Unmanned Aerial Vehicle Using Intelligent Agents for Knowledge Creation. Digita lAvionies Systems Conference, 2004.
[29]唐强,朱志强,王建元.国外无人机自主飞行控制研究[J].系统工程与电子技术,2004.
[30]高正,陈仁良.直升机飞行动力学[M].科学出版社, 2003.
[31] Munzinger C. Development of a Real-Time Flight Simulator for an Experimental Model Helicopter [D]. Master's thesis, Georgia Institute of Technology, School of Aerospace Engineering, Atlanta, GA 30332, July 1997.
[32]杨超,张晓谷.直升机飞行动力学数学模型研究特点分析[J].飞行力学, 1998.
[33] Mettler B. Modeling Small-Scale Unmanned Rotorcraft for Advanced Flight Control Design [D]. Minnesota: Department of Mechanical Engineering Carnegie Mellon University, 2001.
[34]胡寿松.自动控制原理[M].科学出版社,2005.
[35]史忠科.线性系统理论[M].科学出版社,2008.
[36]王俊国.基于神经网络的建模方法与控制策略研究[D].华中科技大学博士论文,2004.
[37]曹梦龙,安世奇.控制系统计算机仿真技术[M].化学工业出版社,2009.
[38] Rysdyk R T, Calise A J, and R. T. N. Chen. Nonlinear Adaptive Control of Tiltrotor AircraftUsing Neural Networks [A]. AIAA/SAE World Aviation Congress [C], Anaheim,CA, 1997.
[39] Rysdyk R T, Calise A J. Adaptive Model Inversion Flight Control for Tiltrotor Aircraft [A]. American Helicopter Society Annual Forum [C], Washington, DC, May 1998.
[40] American Helicopter Society Annual Forum [C], Washington, DC, May 1998.Calise A J, Lee H, Kim N. High Bandwidth Adaptive Flight Control [A]. AIAA Guidance,Navigation and Control Conference [C], Denver, CO, August 2000.
[41]樊京,刘叔军,盖晓华,崔世林. Matlab控制系统应用与实例[M].清华大学出版社,2008,205-230.
[42]杨一栋.直升机飞行控制[M].国防工业出版社, 2007,26-56,94-107.
[43]杨军,吴希明,凡永华,袁博.倾转旋翼机飞行控制[M].航空工业出版社,2006,24-34,64-70.
[44] Hovakimyan N, Kim N, Calise A J, etal. Adaptive Output Feedback for High-Bandwidth Control of an Unmanned Helicopter [A]. AIAA Guidance, Navigation and Control Conference [C], 2001.
[45] Johnson E N, Calise A J, El-Shirbiny H A, etal. Feedback Linearization with Neural Network Augmentation Applied to X-33 Attitude Control [R].AIAA Paper:AIAA-2000-4157.
[46] Johnson E N, Calise A J. Neural Network Adaptive Control of Systems with Input Saturation [A].American Control Conference[C], Arlington, Virginia, June 2001.
[47] Calise A J. Neural Network-Based Adaptive Control for Nonlinear Flight Regimes [R],Technical Report 2002: 1-50.
[48] Karli Wastson,Christian Nagel著,齐立波,黄静译. C#入门经典[M].清华大学出版社,2006,151-220,331-378.
[49] Walter Savitch著,周靖译. C++面向对象程序设计:基础、数据结构与编程思想[M].清华大学出版社,2003.115-218,262-280.
[50]韩立群.人工神经网络教程[M].北京邮电大学出版社,2006,76-120.
[51]徐丽娜.神经网络控制[M].哈尔滨工业大学出版社,1999.116-145.
[52]张洪润.智能技术:系统设计与开发[M].北京航空航天大学出版社,2007.60-95,236-247.
[53] Lee S. Neural Network Based Adaptive Control and its Applications to Aerial Vehicles [D]. Atlanta: Georgia Institute of Technology, 2001.
[54] Johnson E N, Kannan S. Adaptive Flight Control for an Autonomous Unmanned Helicopter [R]. AIAA Paper: AIAA-2002-4439.
[55] Kannan S, Johnson E N. Adaptive Trajectory Based Flight Control for Autonomous Helicopters [A]. AIAA Digital Avionics Conference [C], Irvine, CA, October 2002.
[56] Yong Liu.Neural Adaptive Nonlinear Tracking Using Trajectory Linearization [D].Russ: College of Engineering and Technology Ohio University,2007.