单脉冲PD雷达弱小目标检测算法研究
|
摘要
现代战争中大量使用隐身飞机、反辐射导弹、巡航导弹等飞行器。它们的雷达回波能量比传统飞行器微弱得多，信噪比很低，传统的检测算法难以获得满意的检测性能。通过增大发射功率、提高天线孔径的方法虽然可以改善信噪比，但是具体实现上会受到诸多现实条件的限制，工程实现难度很大，甚至不可行。因此，从频域、时域、空域挖掘目标更多有用信息，并通过先进的信号处理实现弱小信号的检测成为必要的技术途径。
本文针对单脉冲PD雷达弱小目标检测问题，充分挖掘了雷达三个通道中的信息，并综合利用有助于区分目标和噪声的信息(如角误差、复角相位、航迹等)进行了检测判决，有效提高了雷达的检测能力。本文的主要内容包括以下几个方面。绪论部分在介绍课题背景及研究意义的基础上，简要介绍了单脉冲和PD雷达的概念及其工作原理，并阐述了国内外研究现状及研究动态。
第二章为单脉冲PD雷达信号检测的基本方法。在给出单脉冲PD雷达回波信号接收模型的基础上，介绍了该体制下的信号检测典型流程；分析总结了噪声背景下弱小目标检测技术的特点，提出了解决问题的几种可能方法和思路。第三章为差通道信息辅助的弱小目标检测算法。首先，分析了单脉冲体制下测角误差的分布特性，探讨了增大目标信号和噪声角误差分布差异的可能途径；然后，利用角误差分布差异特征构造了检验统计量，并在非相参积累检测算法的基础上，增加了检验统计量辅助检测支路，仿真结果表明新算法在一定程度上提高了非相参积累检测算法的性能。差通道中除了角误差信息外，还存在复角相位信息，这也是目标有用信息之一。因此，本章还设计了以归一化的信号幅度、角误差和复角相位为输入的神经网络目标检测算法。该算法试图综合利用雷达三通道有用信息，通过概率神经网络强大的分类能力达到更好的区分目标和噪声的目的。仿真结果表明该算法与非相参积累检测算法相比具有更优的性能。
第四章为基于DP-TBD的高重频PD雷达检测算法。首先，介绍了目标运动模型和观测模型；然后，阐述了基于动态规划的TBD算法原理及流程，并提出了高重频模式下的DP-TBD算法；最后，结合第三章中的角误差信息，设计了在角误差-多普勒-时间三维空间上的DP-TBD算法，仿真结果表明新算法能进一步提高检测性能。
The flyers, such as stealth aircraft, Anti radiation Missile, and Cruise Missile and so on, are widely used in modern war. The radar echoes of them are much weaker than those of traditional flyers. The signal to noise ratio(SNR) is too low to detect target reliably with general algorithms. Although by increasing the transmitter power, antenna aperture would improve SNR, the concrete realization is always subject to many reality conditions, the project implementation is very difficult, even impossible. Therefore, digging more useful target information from Frequency domain, Time domain and Spatial domain, and through advanced signal processing to detect weak target become necessary technical means
The dissertation focuses on the problem of weak target detection with Monopole Doppler radar. In this paper, we fully use helpful information of three channels of radar (such as the angle error, complex phase angle and track etc.) to distinguish target from noises. New algorithms effectively improve radar’s detection capability. This paper mainly covers the following aspects:
Chapter 1 introduces the background and significance of the problem to be studied, and briefly introduces the concept of Monopulse and PD radar besides their working elements. At last the corresponding techniques and research trends at home and abroad are described.
Chapter 2 is the signal detection basic theory of Monopulse PD radar. Firstly, the signal echo receiving model of Monopulse PD radar is presented, then introduces the typical flow of signal detection under the system. Finally, analyzes and summarizes the characteristics of weak target problem, proposes several possible methods to solve the problem.
Chapter 3 is the new algorithm designing which is based on difference channel information. Firstly, the analysis of angle error measurement under Monopulse system is proposed, and the possible way of increasing the difference of angle distribution between the target and noises is discussed. Secondly, the test statistic based on angle error is constructed, and in the traditional non phase coherent integration algorithm, the test statistic auxiliary detection branch is added. The simulation results show that the new algorithm improves performance of primary non coherent detection algorithm. In addition to angle error information, difference channel also have phase information, which is useful information for detection too. Therefore, dim target detection algorithm based on neural network of which inputs are normalized signal amplitude, angle error and phase of angle is proposed. New algorithm attempts to make fully use of three channel information, and combine the powerful classification ability of probabilistic neural network to better distinguish between the target and noise. Simulation results show that the algorithm has better performance compared with non coherent integration algorithm.
Chapter 4 is Track Before Detect algorithm based on dynamic programming in high PRF PD radar. Firstly, target motion model and observation model are briefly introduced. Secondly, the principles and processes of TBD algorithm based on dynamic programming are described, and DP TBD algorithm in HPRF mode is proposed. At last, with the angle error information a new DP TBD algorithm which process in the Angle Doppler Time three dimensional space is proposed, Simulation results show that the new algorithm can further improve the detection capability.
本文针对单脉冲PD雷达弱小目标检测问题，充分挖掘了雷达三个通道中的信息，并综合利用有助于区分目标和噪声的信息(如角误差、复角相位、航迹等)进行了检测判决，有效提高了雷达的检测能力。本文的主要内容包括以下几个方面。绪论部分在介绍课题背景及研究意义的基础上，简要介绍了单脉冲和PD雷达的概念及其工作原理，并阐述了国内外研究现状及研究动态。
第二章为单脉冲PD雷达信号检测的基本方法。在给出单脉冲PD雷达回波信号接收模型的基础上，介绍了该体制下的信号检测典型流程；分析总结了噪声背景下弱小目标检测技术的特点，提出了解决问题的几种可能方法和思路。第三章为差通道信息辅助的弱小目标检测算法。首先，分析了单脉冲体制下测角误差的分布特性，探讨了增大目标信号和噪声角误差分布差异的可能途径；然后，利用角误差分布差异特征构造了检验统计量，并在非相参积累检测算法的基础上，增加了检验统计量辅助检测支路，仿真结果表明新算法在一定程度上提高了非相参积累检测算法的性能。差通道中除了角误差信息外，还存在复角相位信息，这也是目标有用信息之一。因此，本章还设计了以归一化的信号幅度、角误差和复角相位为输入的神经网络目标检测算法。该算法试图综合利用雷达三通道有用信息，通过概率神经网络强大的分类能力达到更好的区分目标和噪声的目的。仿真结果表明该算法与非相参积累检测算法相比具有更优的性能。
第四章为基于DP-TBD的高重频PD雷达检测算法。首先，介绍了目标运动模型和观测模型；然后，阐述了基于动态规划的TBD算法原理及流程，并提出了高重频模式下的DP-TBD算法；最后，结合第三章中的角误差信息，设计了在角误差-多普勒-时间三维空间上的DP-TBD算法，仿真结果表明新算法能进一步提高检测性能。
The flyers, such as stealth aircraft, Anti radiation Missile, and Cruise Missile and so on, are widely used in modern war. The radar echoes of them are much weaker than those of traditional flyers. The signal to noise ratio(SNR) is too low to detect target reliably with general algorithms. Although by increasing the transmitter power, antenna aperture would improve SNR, the concrete realization is always subject to many reality conditions, the project implementation is very difficult, even impossible. Therefore, digging more useful target information from Frequency domain, Time domain and Spatial domain, and through advanced signal processing to detect weak target become necessary technical means
The dissertation focuses on the problem of weak target detection with Monopole Doppler radar. In this paper, we fully use helpful information of three channels of radar (such as the angle error, complex phase angle and track etc.) to distinguish target from noises. New algorithms effectively improve radar’s detection capability. This paper mainly covers the following aspects:
Chapter 1 introduces the background and significance of the problem to be studied, and briefly introduces the concept of Monopulse and PD radar besides their working elements. At last the corresponding techniques and research trends at home and abroad are described.
Chapter 2 is the signal detection basic theory of Monopulse PD radar. Firstly, the signal echo receiving model of Monopulse PD radar is presented, then introduces the typical flow of signal detection under the system. Finally, analyzes and summarizes the characteristics of weak target problem, proposes several possible methods to solve the problem.
Chapter 3 is the new algorithm designing which is based on difference channel information. Firstly, the analysis of angle error measurement under Monopulse system is proposed, and the possible way of increasing the difference of angle distribution between the target and noises is discussed. Secondly, the test statistic based on angle error is constructed, and in the traditional non phase coherent integration algorithm, the test statistic auxiliary detection branch is added. The simulation results show that the new algorithm improves performance of primary non coherent detection algorithm. In addition to angle error information, difference channel also have phase information, which is useful information for detection too. Therefore, dim target detection algorithm based on neural network of which inputs are normalized signal amplitude, angle error and phase of angle is proposed. New algorithm attempts to make fully use of three channel information, and combine the powerful classification ability of probabilistic neural network to better distinguish between the target and noise. Simulation results show that the algorithm has better performance compared with non coherent integration algorithm.
Chapter 4 is Track Before Detect algorithm based on dynamic programming in high PRF PD radar. Firstly, target motion model and observation model are briefly introduced. Secondly, the principles and processes of TBD algorithm based on dynamic programming are described, and DP TBD algorithm in HPRF mode is proposed. At last, with the angle error information a new DP TBD algorithm which process in the Angle Doppler Time three dimensional space is proposed, Simulation results show that the new algorithm can further improve the detection capability.
引文
[1]胡体玲. 3mm波段高分辨力单脉冲雷达技术研究[M].南京理工大学博士学位论文. 2007.
[2]娄军.机载雷达低截面积高速目标检测跟踪技术研究[D].湖南长沙:国防科技大学, 2008.
[3]孙立宏.雷达弱小目标检测前跟踪算法研究[D],陕西西安:西安电子科技大学, 2007.
[4]王俊.微弱目标信号积累检测的方法研究.陕西西安:西安电子科技大学[D], 1999.
[5]廖云.毫米波单脉冲PD制导雷达抗拖拽式干扰研究[D].湖南长沙:湖南大学, 2009.
[6]郭少南.强杂波下微弱目标检测技术研究[D].四川成都:电子科技大学, 2006.
[7] Skolnik M I. Radar Handbook. New York: McGraw Hill, Inc, 1990.
[8]毛士艺,张瑞生等.脉冲多普勒雷达[M].北京:国防工业出版社,1990.
[9]刘彬,微弱目标检测前跟踪算法研究[D].四川成都:电子科技大学, 2010.
[10]何友等.雷达自动检测与恒虚警处理[M].北京:清华大学出版社, 1999.
[11]何友等.雷达自动检测和CFAR处理方法综述[J].系统工程与电子科技, 2001.
[12] Ye zhenru, Zhu Zhaoda, Huang Xingping. CFAR Analysis for Medium PRF Airbone Pulse Doopler Radars. IEEE National Aerospace and Electronics Conference, 1988: 259~263.
[13] Ringel M B, Mooney D H, Long W H. F 16 Pulse Doppler Radar(AN/APG 16) Performance. IEEE Trans. On AES, 1983, 19(1):147~157.
[14] Maggi M, Pistoia D. False Alarm Probability for a Doppler Detector with a Threshold Adaptive to the Spectral Characteristics of the Signal. Proc. Of CIE 1991 International Conference on Radar, 1991:427~430.
[15] Trunk G V, Gordon G B, Cantrell H. False Alarm Control Using Doppler. IEEE Trans. on AES, 1990, 17(4):501~510.
[16] Kennydy M P, Chua L O. Neural networks for nonlinear programming. IEEE Trans. Circuits and Systems, 35(5),554 562(1988).
[17]阎平凡,张长水.人工神经网络与模拟进化计算[M].北京:清华大学出版社,2005.
[18]周辉.基于神经网络的高分辨雷达目标检测研究[D].湖南长沙:长沙理工大学, 2010.
[19] MICHAEL W ROTH Neural Networks for Extraction of Weak Targets in High Clutter Environments. IEEE, 1989, 1210 1216.
[20] Leung H, Dubash N, Xie N. Detection of small Objects in Clutter Using a GA2RBF Netural Network. IEEE Trans on AES, 2002, 38(1).
[21]杨恒,张贤达.模糊神经网络分类器[J].信号处理, 1996.
[22]温晓君.海杂波背景下基于神经网络的目标检测[J].系统仿真学报, 2007, 19(7):1639~1641.
[23]赖作镁,王敬儒.基于RBF神经网络的复杂背景下的运动目标检测[J]. 2007, 34(2): 250~252.
[24]吴顺君,梅晓春.雷达信号处理和数据处理技术[M].北京:电子工业出版社, 2008.
[25] Carlson B D, Evans E D, Wilson S L. Search radar detection and track with the Hough transform, Part I, II, III: System concept. IEEE Transaction On Aerospace and Electronic Systems, 1994, 30(I): 102—125.
[26] Moqiseh A, Nayebi M M. 3 D Hough Transform For Surveillance Radar Target Detection. Proc. of the 2008 IEEE Radar conference, 2008:882~886.
[27]王国宏,苏峰,毛士艺,何友.杂波环境下基于Hough变换和逻辑的快速航迹起始算法[J].系统仿真学报, 2002, 14(7): 874—875.
[28]王峰,罗利强,郝小宁.航迹起始算法及其性能[J].仿真火控雷达技术, 2009, 38(1): 56 58.
[29] Gordon N J, Salmond D J, Smith A F M, Novel approach to nonlinear/ non Gaussian Bayesian state estimation[J]. IEEE PROCEEDING, 1993, 140(2).
[30] Rutten M G, Gordon N J, Maskell S. Particle Based Track Before Detect in Rayleigh Noise. Signal and Data Processing of Small Targets, Proceedings of SPIE, 2004, 5428:509 519.
[31] Brekke E, Kirubarajan T, Tharmarasa R. Tracking dim targets using integrated clutter estimation. Signal and Tata Processing of Small Targets, Proceedings of SPIE, 2007.
[32] Davey S J, Cheung B, Rutten M G. Track Before Detect for Sensors with Complex Measurements. Information Fusion, 2009, 618 625.
[33] Su H T, Wu T P. Rao—Blackwellised particle filter based track before detect algorithm. Signal Processing, IET, 2008, 2(2):169 176.
[34] Barniv Y. Dynmic Programming Solution for Detecting Dim Moving Targets. IEEE Transaction on Aerospace Electronics Systems, 1985, 21(1): 144~156.
[35] Jennifer L. Hartnon. Track Before Detect performance for a high PRF search mode[J], 1991.
[36] Wallace W R. The use of Track Before Detect in Pulse Doppler radar. Radar, 2002: 315 319.
[37] Johnston, L.A, Krishnamurthy, V. Performance analysis of a dynamic programming track before detect algorithm. Aerospace and Electronic Systems, IEEE Transactions on , Jan 2002, pp.228 242.
[38] Johnston, L.A., Krishnamurthy, V. Performance analysis of a track before detect dynamic programming algorithm. Acoustics, Speech, and Signal Processing, 2000. pp.49 52.
[39]孙立宏,王俊.高速运动雷达弱小目标检测方法研究[M].系统工程与电子科学, 2008,30(2): 257 4.
[40]宋慧波,高梅国,田黎育.一种有效的雷达微弱目标检测算法[J].仪器仪表学报, 2006, 27(6): 1326 1327, 1339.
[41]范红旗.主动寻的制导中机动目标运动模式辨识技术[D].湖南长沙:国防科学技术大学, 2008.
[42] MALLAT S, HWANG W L. Singularity detection and processing with wavelets[J]. IEEE Transaction on Information Theory, 1992,38(2): 617 643.
[43] DONOHO D L, JOHNSTONE I M. Threshold selection for wavelet shrinkage of noisy data[J]. Proceedings of the 16th Annual International Conference of the IEEE, 1994, 4:794 795.
[44]王祖林,张孟,段世忠,周荫清.比相单脉冲雷达测角与角闪烁研究[J].航空学报. 2001,22:S26 04.
[45] Mark A. Richards. Fundamentals of Radar Signal Processing[M]. McGraw Hill Companies, Inc. 2005
[46]刘清宇,方世良,徐江.联合检测估计及其性能分析[J].声学技术. 2009:28(5): 655~659.
[47] Middleton D, Esposito R. Simultaneous optimum detection and estimation of signals in noise[J]. IEEE Trans. Inform. Theory, 1968, 1T 14(3):434 444.
[48] Hughes E J. Radar Cross Section Modeling Using Genetic Algorithms[D]. Shrivenham: Royal Military College of Science, 1998.
[49] Specht D F. Probabilistic Neural Network for Classification, Mapping or Associative Memory[J]. IEEE International Conference, 1988: 525 532.
[50]库亮.基于动态规划的弱小多目标检测与跟踪[D].江苏无锡:江南大学, 2009.
[2]娄军.机载雷达低截面积高速目标检测跟踪技术研究[D].湖南长沙:国防科技大学, 2008.
[3]孙立宏.雷达弱小目标检测前跟踪算法研究[D],陕西西安:西安电子科技大学, 2007.
[4]王俊.微弱目标信号积累检测的方法研究.陕西西安:西安电子科技大学[D], 1999.
[5]廖云.毫米波单脉冲PD制导雷达抗拖拽式干扰研究[D].湖南长沙:湖南大学, 2009.
[6]郭少南.强杂波下微弱目标检测技术研究[D].四川成都:电子科技大学, 2006.
[7] Skolnik M I. Radar Handbook. New York: McGraw Hill, Inc, 1990.
[8]毛士艺,张瑞生等.脉冲多普勒雷达[M].北京:国防工业出版社,1990.
[9]刘彬,微弱目标检测前跟踪算法研究[D].四川成都:电子科技大学, 2010.
[10]何友等.雷达自动检测与恒虚警处理[M].北京:清华大学出版社, 1999.
[11]何友等.雷达自动检测和CFAR处理方法综述[J].系统工程与电子科技, 2001.
[12] Ye zhenru, Zhu Zhaoda, Huang Xingping. CFAR Analysis for Medium PRF Airbone Pulse Doopler Radars. IEEE National Aerospace and Electronics Conference, 1988: 259~263.
[13] Ringel M B, Mooney D H, Long W H. F 16 Pulse Doppler Radar(AN/APG 16) Performance. IEEE Trans. On AES, 1983, 19(1):147~157.
[14] Maggi M, Pistoia D. False Alarm Probability for a Doppler Detector with a Threshold Adaptive to the Spectral Characteristics of the Signal. Proc. Of CIE 1991 International Conference on Radar, 1991:427~430.
[15] Trunk G V, Gordon G B, Cantrell H. False Alarm Control Using Doppler. IEEE Trans. on AES, 1990, 17(4):501~510.
[16] Kennydy M P, Chua L O. Neural networks for nonlinear programming. IEEE Trans. Circuits and Systems, 35(5),554 562(1988).
[17]阎平凡,张长水.人工神经网络与模拟进化计算[M].北京:清华大学出版社,2005.
[18]周辉.基于神经网络的高分辨雷达目标检测研究[D].湖南长沙:长沙理工大学, 2010.
[19] MICHAEL W ROTH Neural Networks for Extraction of Weak Targets in High Clutter Environments. IEEE, 1989, 1210 1216.
[20] Leung H, Dubash N, Xie N. Detection of small Objects in Clutter Using a GA2RBF Netural Network. IEEE Trans on AES, 2002, 38(1).
[21]杨恒,张贤达.模糊神经网络分类器[J].信号处理, 1996.
[22]温晓君.海杂波背景下基于神经网络的目标检测[J].系统仿真学报, 2007, 19(7):1639~1641.
[23]赖作镁,王敬儒.基于RBF神经网络的复杂背景下的运动目标检测[J]. 2007, 34(2): 250~252.
[24]吴顺君,梅晓春.雷达信号处理和数据处理技术[M].北京:电子工业出版社, 2008.
[25] Carlson B D, Evans E D, Wilson S L. Search radar detection and track with the Hough transform, Part I, II, III: System concept. IEEE Transaction On Aerospace and Electronic Systems, 1994, 30(I): 102—125.
[26] Moqiseh A, Nayebi M M. 3 D Hough Transform For Surveillance Radar Target Detection. Proc. of the 2008 IEEE Radar conference, 2008:882~886.
[27]王国宏,苏峰,毛士艺,何友.杂波环境下基于Hough变换和逻辑的快速航迹起始算法[J].系统仿真学报, 2002, 14(7): 874—875.
[28]王峰,罗利强,郝小宁.航迹起始算法及其性能[J].仿真火控雷达技术, 2009, 38(1): 56 58.
[29] Gordon N J, Salmond D J, Smith A F M, Novel approach to nonlinear/ non Gaussian Bayesian state estimation[J]. IEEE PROCEEDING, 1993, 140(2).
[30] Rutten M G, Gordon N J, Maskell S. Particle Based Track Before Detect in Rayleigh Noise. Signal and Data Processing of Small Targets, Proceedings of SPIE, 2004, 5428:509 519.
[31] Brekke E, Kirubarajan T, Tharmarasa R. Tracking dim targets using integrated clutter estimation. Signal and Tata Processing of Small Targets, Proceedings of SPIE, 2007.
[32] Davey S J, Cheung B, Rutten M G. Track Before Detect for Sensors with Complex Measurements. Information Fusion, 2009, 618 625.
[33] Su H T, Wu T P. Rao—Blackwellised particle filter based track before detect algorithm. Signal Processing, IET, 2008, 2(2):169 176.
[34] Barniv Y. Dynmic Programming Solution for Detecting Dim Moving Targets. IEEE Transaction on Aerospace Electronics Systems, 1985, 21(1): 144~156.
[35] Jennifer L. Hartnon. Track Before Detect performance for a high PRF search mode[J], 1991.
[36] Wallace W R. The use of Track Before Detect in Pulse Doppler radar. Radar, 2002: 315 319.
[37] Johnston, L.A, Krishnamurthy, V. Performance analysis of a dynamic programming track before detect algorithm. Aerospace and Electronic Systems, IEEE Transactions on , Jan 2002, pp.228 242.
[38] Johnston, L.A., Krishnamurthy, V. Performance analysis of a track before detect dynamic programming algorithm. Acoustics, Speech, and Signal Processing, 2000. pp.49 52.
[39]孙立宏,王俊.高速运动雷达弱小目标检测方法研究[M].系统工程与电子科学, 2008,30(2): 257 4.
[40]宋慧波,高梅国,田黎育.一种有效的雷达微弱目标检测算法[J].仪器仪表学报, 2006, 27(6): 1326 1327, 1339.
[41]范红旗.主动寻的制导中机动目标运动模式辨识技术[D].湖南长沙:国防科学技术大学, 2008.
[42] MALLAT S, HWANG W L. Singularity detection and processing with wavelets[J]. IEEE Transaction on Information Theory, 1992,38(2): 617 643.
[43] DONOHO D L, JOHNSTONE I M. Threshold selection for wavelet shrinkage of noisy data[J]. Proceedings of the 16th Annual International Conference of the IEEE, 1994, 4:794 795.
[44]王祖林,张孟,段世忠,周荫清.比相单脉冲雷达测角与角闪烁研究[J].航空学报. 2001,22:S26 04.
[45] Mark A. Richards. Fundamentals of Radar Signal Processing[M]. McGraw Hill Companies, Inc. 2005
[46]刘清宇,方世良,徐江.联合检测估计及其性能分析[J].声学技术. 2009:28(5): 655~659.
[47] Middleton D, Esposito R. Simultaneous optimum detection and estimation of signals in noise[J]. IEEE Trans. Inform. Theory, 1968, 1T 14(3):434 444.
[48] Hughes E J. Radar Cross Section Modeling Using Genetic Algorithms[D]. Shrivenham: Royal Military College of Science, 1998.
[49] Specht D F. Probabilistic Neural Network for Classification, Mapping or Associative Memory[J]. IEEE International Conference, 1988: 525 532.
[50]库亮.基于动态规划的弱小多目标检测与跟踪[D].江苏无锡:江南大学, 2009.