状态自适应无迹卡尔曼滤波算法及其在水下机动目标跟踪中的应用
|
摘要
为了满足水下对抗对机动目标实时跟踪和目标航速、航向准确估计的要求,针对观测量为距离和方位的机动目标跟踪,对传统无迹卡尔曼滤波(UKF)跟踪算法进行了改善。提出根据UKF算法预测值和观测值残差的概率分布自适应调整目标状态噪声方法,使得UKF跟踪算法能够根据目标运动状态及时调整状态方程,在目标机动时减小对预测值的依赖,在目标非机动时增大对预测值的依赖。这种在线实时估计系统噪声状态的跟踪方法更加适用于机动目标的跟踪。数值仿真结果表明:该算法不仅在目标机动时具有良好的跟踪效果,而且在目标非机动时具有准确的估计性能。通过声纳信息综合处理系统验证了状态自适应UKF跟踪算法的性能。
In order to meet the needs for tracking the underwater maneuvering target in real-time and predicting its location in the underwater confrontation environment,the higher requirements are put forward for accurate and fast estimation of target's sailing speed and course. The traditional unscented Kalman filter( UKF) tracking algorithm is improved for maneuvering target tracking with range and azimuth. The improved tracking algorithm can also be used to estimate the system status noise online in real-time without determining the state equation and the state noise variance in advance,thus tracking the maneuvering target adaptively. A novel adaptive UKF method is proposed,which adaptively adjusts the target state noise with residual probability distribution according to the residuals of predicted and observed values of UKF algorithm,so that the UKF tracking algorithm can adjust the state according to the target state the equation reduces the reliance on the predicted value when the target is maneuvering and increases the reliance on the prediction when the target is maneuvering. Primary numerical simulation results show that the algorithm not only has good tracking performance in target maneuvering but also has accurate estimation in non-maneuvering. At last,the tracking performance of state adaptive UKF algorithm is illustrated by sonar simulation system.
In order to meet the needs for tracking the underwater maneuvering target in real-time and predicting its location in the underwater confrontation environment,the higher requirements are put forward for accurate and fast estimation of target's sailing speed and course. The traditional unscented Kalman filter( UKF) tracking algorithm is improved for maneuvering target tracking with range and azimuth. The improved tracking algorithm can also be used to estimate the system status noise online in real-time without determining the state equation and the state noise variance in advance,thus tracking the maneuvering target adaptively. A novel adaptive UKF method is proposed,which adaptively adjusts the target state noise with residual probability distribution according to the residuals of predicted and observed values of UKF algorithm,so that the UKF tracking algorithm can adjust the state according to the target state the equation reduces the reliance on the predicted value when the target is maneuvering and increases the reliance on the prediction when the target is maneuvering. Primary numerical simulation results show that the algorithm not only has good tracking performance in target maneuvering but also has accurate estimation in non-maneuvering. At last,the tracking performance of state adaptive UKF algorithm is illustrated by sonar simulation system.
引文
[1] ANDERSON B D O,MOORE J B,ESLAMI M. Optimal filtering[M]. Englewood Cliffs,NJ,US:Prentice-Hall,1979.
[2]孟祥萍,张本法,苑全德.自适应扩展卡尔曼滤波在移动机器人定位中的应用[J].计算机系统应用,2015,24(12):176-181.MENG X P,ZHANG B F,YUAN Q D. Adaptive extended Kal-man filter in the application of the mobile robot localization[J].Computer Systems&Applications,2015,24(12):176-181.(inChinese)
[3]吴林煌,苏凯雄,郭里婷,等.基于自适应扩展卡尔曼滤波与神经网络的HPA预失真算法[J].自动化学报,2016,42(1):122-130.WU L H,SU K X,GUO L T,et al. HPA predistortion algorithmbased on adaptive extended Kalman filter and neural network[J].Acta Automatic Sinica,2016,42(1):122-130.(in Chinese)
[4] LARA I B. Theory and applications of Monte Carlo simulations[J]. Journal of Socioeconomic Engineering,2015(2):45-48
[5] GORDON N J,SALMOND D J,SMITH A F M. Novel approachto nonlinear/non-Gaussian Bayesian state estimation[J]. IEE Pro-ceedings-F:Radar and Signal Processing,1993,140(2):107-113.
[6] WAN E A,VAN DER MERWE R. The unscented Kalman filterfor nonlinear estimation[C]∥Proceedings of IEEE 2000 AdaptiveSystems for Signal Processing,Communications,and Control Sym-posium. Lake Louise,Alberts,Canada:IEEE,2000:153-158.
[7] JULIER S J,UHLMANN J K. New extension of the Kalman filterto nonlinear systems[J]. Proceedings of SPIE,Signal ProcessingSensor Fusion&Target Recognition VI,1997,3068:182-193.
[8] UHLMANN J K. Algorithm for multiple target tracking[J]. Amer-ican Science,1992,80(2):128-141.
[9]万莉,刘焰春,皮亦鸣. EKF、UKF、PF目标跟踪性能的比较[J].雷达科学与技术,2007,5(1):13-16.WAN L,LIU Y C,PI Y M. Comparing of target tracking perform-ances of EKF,UKF and PF[J]. Radar Science and Technology,2007,5(1):13-16.(in Chinese)
[10]赵治国,朱强,周良杰,等.分布式驱动HEV自适应无迹卡尔曼车速估计[J].中国科学,2016,46(5):481-492.ZHAO Z G,ZHU Q,ZHOU L J,et al. Vehicle speed estimationin driving case based on distributed self-adaptive unscented Kal-man filter for 4WD hybrid electric car[J]. Scientia Sinica Tech-nologica,2016,46(5):481-492.(in Chinese)
[11]侯江宽,马珺,贾华宇.改进型自适应无迹卡尔曼姿态算法[J].传感技术学报,2017,30(10):1518-1526.HOU J K,MA J,JIA H Y. An improved adaptive unsentedKalman filter for attitude determination[J]. Chinese Journal ofSensors and Actuators,2017,30(10):1518-1526.(in Chi-nese)
[12]石勇,韩崇昭.自适应UKF算法在目标跟踪中的应用[J].自动化学报,2011,37(6):755-759.SHI Y,HAN C Z. Adaptive UKF method with applications totarget tracking[J]. Acta Automatic Sinica,2011,37(6):755-759.(in Chinese)
[13]周卫东,乔相伟,吉宇人,等.基于新息和残差的自适应UKF算法[J].宇航学报,2010,31(7):1798-1804.ZHOU W D,QIAO X W,JI Y R,et al. An innovation and re-sidual-based adaptive UKF algorithm[J]. Journal of Astronau-tics,2010,31(7):1798-1804.(in Chinese)
[14]陈政,刘康.闪烁噪声下的自适应无迹卡尔曼算法[J].软件,2018,39(1):70-74.CHEN Z,LIU K. Adaptive unsented Kalman algorithm for flickernoise[J]. Computer Engineering&Software,2018,39(1):70-74.(in Chinese)
[2]孟祥萍,张本法,苑全德.自适应扩展卡尔曼滤波在移动机器人定位中的应用[J].计算机系统应用,2015,24(12):176-181.MENG X P,ZHANG B F,YUAN Q D. Adaptive extended Kal-man filter in the application of the mobile robot localization[J].Computer Systems&Applications,2015,24(12):176-181.(inChinese)
[3]吴林煌,苏凯雄,郭里婷,等.基于自适应扩展卡尔曼滤波与神经网络的HPA预失真算法[J].自动化学报,2016,42(1):122-130.WU L H,SU K X,GUO L T,et al. HPA predistortion algorithmbased on adaptive extended Kalman filter and neural network[J].Acta Automatic Sinica,2016,42(1):122-130.(in Chinese)
[4] LARA I B. Theory and applications of Monte Carlo simulations[J]. Journal of Socioeconomic Engineering,2015(2):45-48
[5] GORDON N J,SALMOND D J,SMITH A F M. Novel approachto nonlinear/non-Gaussian Bayesian state estimation[J]. IEE Pro-ceedings-F:Radar and Signal Processing,1993,140(2):107-113.
[6] WAN E A,VAN DER MERWE R. The unscented Kalman filterfor nonlinear estimation[C]∥Proceedings of IEEE 2000 AdaptiveSystems for Signal Processing,Communications,and Control Sym-posium. Lake Louise,Alberts,Canada:IEEE,2000:153-158.
[7] JULIER S J,UHLMANN J K. New extension of the Kalman filterto nonlinear systems[J]. Proceedings of SPIE,Signal ProcessingSensor Fusion&Target Recognition VI,1997,3068:182-193.
[8] UHLMANN J K. Algorithm for multiple target tracking[J]. Amer-ican Science,1992,80(2):128-141.
[9]万莉,刘焰春,皮亦鸣. EKF、UKF、PF目标跟踪性能的比较[J].雷达科学与技术,2007,5(1):13-16.WAN L,LIU Y C,PI Y M. Comparing of target tracking perform-ances of EKF,UKF and PF[J]. Radar Science and Technology,2007,5(1):13-16.(in Chinese)
[10]赵治国,朱强,周良杰,等.分布式驱动HEV自适应无迹卡尔曼车速估计[J].中国科学,2016,46(5):481-492.ZHAO Z G,ZHU Q,ZHOU L J,et al. Vehicle speed estimationin driving case based on distributed self-adaptive unscented Kal-man filter for 4WD hybrid electric car[J]. Scientia Sinica Tech-nologica,2016,46(5):481-492.(in Chinese)
[11]侯江宽,马珺,贾华宇.改进型自适应无迹卡尔曼姿态算法[J].传感技术学报,2017,30(10):1518-1526.HOU J K,MA J,JIA H Y. An improved adaptive unsentedKalman filter for attitude determination[J]. Chinese Journal ofSensors and Actuators,2017,30(10):1518-1526.(in Chi-nese)
[12]石勇,韩崇昭.自适应UKF算法在目标跟踪中的应用[J].自动化学报,2011,37(6):755-759.SHI Y,HAN C Z. Adaptive UKF method with applications totarget tracking[J]. Acta Automatic Sinica,2011,37(6):755-759.(in Chinese)
[13]周卫东,乔相伟,吉宇人,等.基于新息和残差的自适应UKF算法[J].宇航学报,2010,31(7):1798-1804.ZHOU W D,QIAO X W,JI Y R,et al. An innovation and re-sidual-based adaptive UKF algorithm[J]. Journal of Astronau-tics,2010,31(7):1798-1804.(in Chinese)
[14]陈政,刘康.闪烁噪声下的自适应无迹卡尔曼算法[J].软件,2018,39(1):70-74.CHEN Z,LIU K. Adaptive unsented Kalman algorithm for flickernoise[J]. Computer Engineering&Software,2018,39(1):70-74.(in Chinese)