Multi-try and Multi-model Particle Filter for Maneuvering Target Tracking
-
摘要: 针对机动目标跟踪问题,提出了一种多点测试多模型粒子滤波算法(Independence multi-try method, IMTM).整个算法分为两个阶段,第一阶段为利用多点测试(Multi-try method, MTM)结构从各模型产生的粒子中选取一个最优粒子,实现了模型间的交互;第二阶段为利用IMH (Independence Metropolis-Hastings)滤波算法对第一阶段产生的粒子进行取舍,完成整个状态估计.相对于传统的交互式多模型(Interacting multiple model, IMM)算法,该算法无需事先设定模型转移概率 矩阵且为整体并行结构,结构简单,能够充分地交互各模型之间的粒子,进而自动有效地调整各模型权值比重,降低了人为干扰.仿真表明,该算法能够有效地降低滤波峰值误差,整体跟踪精度较高,算法的实时性较好.Abstract: A novel multiple try and multiple model particle filter named independence multi-try method (IMTM) is presented for maneuvering target tracking. It can be divided into two stages: 1) A multi-try test structure is utilized to choose an optimal particle from a set of particles generated by multiple models; this stage brings about model interaction. 2) Independence Metropolis-Hastings (IMH) will reject or accept the particle produced in stage 1. This method does not require the model conversion matrix so that it can reduce the man-made error introduced by the model conversion matrix as used in traditional methods. It is a parallel and simple structure particle filter. With this method, the particles are exchanged between models automatically and effectively to adjust the model weight. Simulations show that this algorithm can effectively reduce the peak-error, and is a real-time algorithm.
-
[1] Blom H A P, Bar-shalom Y. The interacting multiple model algorithm for systems with Markovian switching coefficients. IEEE Transactions on Automatic Control, 1988, 33(8): 780 -783 [2] [2] Li X R, Jilkov V P. Survey of maneuvering target tracking, Part V. Multiple-model methods. IEEE Transactions on Aerospace and Electronic Systems, 2005, 41(4): 1255-1321 [3] Shi Yong, Han Chong-Zhao. Adaptive UKF method with applications to target tracking. Acta Automatica Sinica, 2011, 37(6): 755-759(石勇, 韩崇昭. 自适应UKF算法在目标跟踪中的应用. 自动化学报, 2011, 37(6): 755-759) [4] [4] Doucet A, Johansen A M. A tutorial on particle filtering and smoothing: fifteen years later. Handbook of Nonlinear Filtering, 2009, 12: 656-704 [5] [5] Foo P H, Ng G W. Combining the interacting multiple model method with particle filters for manoeuvring target tracking. IET Radar, Sonar and Navigation, 2011, 5(3): 234 -255 [6] [6] Gordon N J, Maskell S, Kirubarajan T. Efficient particle filters for joint tracking and classification. In: Proceedings of the 2002 SPIE 4728: Signal and Data Processing of Small Targets. Orlando, FL, USA: SPIE, 2002. 439-449 [7] [7] Hong S H, Shi Z G, Chen K S. Easy-hardware-implementation MMPF for maneuvering target tracking: algorithm and architecture. Journal of Signal Processing Systems, 2010, 61(3): 259-269 [8] [8] McGinnity S, Irwin G W. Multiple model bootstrap filter for maneuvering target tracking. IEEE Transactions on Aerospace and Electronic Systems, 2000, 36(3): 1006-1012 [9] [9] Yang N, Tian W F, Jin Z H. An interacting multiple model particle filter for manoeuvring target location. Measurement Science and Technology, 2006, 17(6): 1307-1311 [10] Boers Y, Driessen J N. Interacting multiple model particle filter. IEE Proceedings Radar, Sonar, and Navigation, 2003, 150(5): 344-349 [11] Sankaranarayanan A C, Chellappa R, Srivastava A. Algorithmic and architectural design methodology for particle filters in hardware. In: Proceedings of the 2005 IEEE International Conference on Computer Design: VLSI in Computers and Processors. San Jose, CA, USA: IEEE, 2005. 275- 280 [12] Miao L F, Zhang J J, Chakrabarti C, Papandreou-Suppappola A. A new parallel implementation for particle filters and its application to adaptive waveform design. In: Proceedings of the 2010 IEEE Workshop on Signal Processing Systems (SIPS). San Francisco, CA: IEEE, 2010. 978(1): 19-24 [13] Liu J S. Monte Carlo Strategies in Scientific Computing. New York: Springer Science and Business Media, 2004. 111 -123 [14] Liu J S, Liang F M, Wong W H. The multiple-try method and local optimization in Metropolis sampling. Journal of the American Statistical Association, 2000, 95(449): 121- 134 [15] Qin Z S, Liu J S. Multi-point Metropolis method with application to hybrid Monte Carlo. Journal of Computational Physics, 2001, 172(2): 827-840 [16] Storvik G. On the flexibility of Metropolis-Hastings acceptance probabilities in auxiliary variable proposal generation. Scandinavian Journal of Statistics, 2011, 38(2): 342-358 [17] Pandolfi S, Bartolucci F, Friel N. A generalization of the multiple-try Metropolis algorithm for Bayesian estimation and model selection. In: Proceedings of the 13th International Conference on Artificial Intelligence and Statistics. 2010. 581-588 [18] Casarin R, Craiu R, Leisen F. Interacting multiple try algorithms with different proposal distributions. Statistics and Computing, 2013, 23(2): 185-200 [19] Blom H A P, Bloem E A. Exact Bayesian and particle filtering of stochastic hybrid systems. IEEE Transactions on Aerospace and Electronic Systems, 2007, 43(1): 55-70 [20] Sankaranarayanan A C, Srivastava A, Chellappa R. Algorithmic and architectural optimizations for computationally efficient particle filtering. IEEE Transactions on Image Processing, 2008, 17(5): 737-748 [21] Arulampalam M S, Maskell S, Gordon N, Clapp T. A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking. IEEE Transactions on Signal Processing, 2002, 50(2): 174-188 [22] Linder S P, Schell C. A non-Bayesian segmenting tracker for highly maneuvering targets. IEEE Transactions on Aerospace and Electronic Systems, 2005, 41(4): 1168-1177 [23] Wang Xiao, Han Chong-Zhao. A multiple model particle filter for maneuvering target tracking based on composite sampling. Acta Automatica Sinica, 2013, 39(7): 1152-1156(王晓, 韩崇昭. 基于混合采样的多模型机动目标跟踪算法. 自动化学报, 2013, 39(7): 1152-1156)
点击查看大图
计量
- 文章访问数: 2232
- HTML全文浏览量: 99
- PDF下载量: 1351
- 被引次数: 0