未知量测噪声分布下的CBMeMBer算法
|
摘要
针对目标跟踪过程中量测噪声概率分布等先验知识无法准确获取的问题,提出一种基于风险评估的势均衡多目标多伯努利(RE-CBMeMBer)滤波算法。采用CBMeMBer算法的序贯蒙特卡洛实现,在粒子预测后利用风险函数和评估函数计算粒子风险值,并用评估结果更新粒子权值。避免了计算似然函数且不依赖量测噪声的概率分布。仿真表明:与SMC-CBMeMBer算法相比,RE-CBMeMBer算法具有更好的实时性,特别是当量测噪声分布未知时,具有更高的跟踪精度和稳定性。
Aiming at the problem of probabilistic distribution of measurement noise and so on are unable to be acquired accurately,a Risk Evaluation-based Cardinality Balanced Multi-Target Multi-Bernoulli(RE-CBMeMBer)filter algorithm is proposed. The algorithm adopts the Sequential Monte Carlo implementation of CBMeMBer algorithm,calculates the particle risk value using the risk function and evaluation function after particle prediction,and updates the particle weight value with the evaluation result. The algorithm avoids the likelihood function calculation and does not depend on the probabilistic distribution of measurement noise. The simulation results show that RE-CBMeMBer filter algorithm has better real-time performance in comparison with SMC-CBMeMBer filter algorithm,especially the algorithm has higher robustness and stability in unknown measurement noise distribution environment.
Aiming at the problem of probabilistic distribution of measurement noise and so on are unable to be acquired accurately,a Risk Evaluation-based Cardinality Balanced Multi-Target Multi-Bernoulli(RE-CBMeMBer)filter algorithm is proposed. The algorithm adopts the Sequential Monte Carlo implementation of CBMeMBer algorithm,calculates the particle risk value using the risk function and evaluation function after particle prediction,and updates the particle weight value with the evaluation result. The algorithm avoids the likelihood function calculation and does not depend on the probabilistic distribution of measurement noise. The simulation results show that RE-CBMeMBer filter algorithm has better real-time performance in comparison with SMC-CBMeMBer filter algorithm,especially the algorithm has higher robustness and stability in unknown measurement noise distribution environment.
引文
[1]BAR-SHALOM Y,FORTMANN T E.Tracking and data association[M].San Diego,CA:Academic,1988:1-34.
[2]VO B N,MA W K.The Gaussian mixture probability hypothesis density filter[J].IEEE Transactions on Signal Processing,2006,54(11):4091-4104.
[3]VO B N,SINGH S,DOUCET A.Sequential Monte Carlo methods for multi-target filtering with random finite sets[J].IEEE Transactions on Aerospace and Electronic Systems,2005,41(4):1224-1245.
[4]UKMKE M,ERDINC O,WILLETT P.GMTI tracking via the Gaussian mixture cardinalized probability hypothesis density filter[J].IEEE Transactions on Aerospace and Electronic Systems,2010,46(4):1821-1833.
[5]MAHLER R.Statistical multisource-multitarget information fusion[M].London:Artech House,2007.
[6]VO B T,VO B N,CANTONI A.The cardinality balanced multi-target multi-bernoulli filter and its implementations[J].IEEE Transactions on Signal Processing,2009,57(2):409-423.
[7]周承兴,刘贵喜.未知测量噪声分布下的多目标跟踪算法[J].航空学报,2010,31(11):2228-2237.
[8]李翠芸,王荣,姬红兵.基于变分贝叶斯势均衡多目标多伯努利滤波的多扩展目标跟踪算法[J].控制理论与应用,2015,32(2):187-195.
[9]赵斌,胡建旺,吉兵.多机动目标跟踪的BFG-GMPHD算法[J].火力与指挥控制,2016,41(11):159-162.
[10]赵凯,胡建旺,吉兵.基于H∞滤波的无序量测更新算法[J].火力与指挥控制,2017,42(2):159-162.
[11]MIGUEZ J,BUGALLO M F,DJURIC P M.A new class of particle filters for random dynamic systems with unknown statistics[J].EURASIP Journal on Advances Signal Processing,2004,2004(15):2278-2294.
[12]MIGUEZ J,XU S S,Bugallo M F,et al.A novel particle filtering approach and its application to target tracking[C]//Proceedings of IEEE Aerospace Conference.2004:1886-1894.
[13]MIGUEZ J.Analysis of a sequential Monte Carlo method for optimization in dynamical systems[J].Signal Processing,2010,90(5):1609-1622.
[2]VO B N,MA W K.The Gaussian mixture probability hypothesis density filter[J].IEEE Transactions on Signal Processing,2006,54(11):4091-4104.
[3]VO B N,SINGH S,DOUCET A.Sequential Monte Carlo methods for multi-target filtering with random finite sets[J].IEEE Transactions on Aerospace and Electronic Systems,2005,41(4):1224-1245.
[4]UKMKE M,ERDINC O,WILLETT P.GMTI tracking via the Gaussian mixture cardinalized probability hypothesis density filter[J].IEEE Transactions on Aerospace and Electronic Systems,2010,46(4):1821-1833.
[5]MAHLER R.Statistical multisource-multitarget information fusion[M].London:Artech House,2007.
[6]VO B T,VO B N,CANTONI A.The cardinality balanced multi-target multi-bernoulli filter and its implementations[J].IEEE Transactions on Signal Processing,2009,57(2):409-423.
[7]周承兴,刘贵喜.未知测量噪声分布下的多目标跟踪算法[J].航空学报,2010,31(11):2228-2237.
[8]李翠芸,王荣,姬红兵.基于变分贝叶斯势均衡多目标多伯努利滤波的多扩展目标跟踪算法[J].控制理论与应用,2015,32(2):187-195.
[9]赵斌,胡建旺,吉兵.多机动目标跟踪的BFG-GMPHD算法[J].火力与指挥控制,2016,41(11):159-162.
[10]赵凯,胡建旺,吉兵.基于H∞滤波的无序量测更新算法[J].火力与指挥控制,2017,42(2):159-162.
[11]MIGUEZ J,BUGALLO M F,DJURIC P M.A new class of particle filters for random dynamic systems with unknown statistics[J].EURASIP Journal on Advances Signal Processing,2004,2004(15):2278-2294.
[12]MIGUEZ J,XU S S,Bugallo M F,et al.A novel particle filtering approach and its application to target tracking[C]//Proceedings of IEEE Aerospace Conference.2004:1886-1894.
[13]MIGUEZ J.Analysis of a sequential Monte Carlo method for optimization in dynamical systems[J].Signal Processing,2010,90(5):1609-1622.