Monte Carlo-based filter for target tracking with feature measurement

D. Angelova1, B. Vassileva1, Tz. Semerdjiev1
1Central Laboratory for Parallel Processing, Bulgarian Academy of Sciences, Sofia, Bulgaria

Tóm tắt

Monte Carlo-based algorithm for tracking maneuvering target with a feature measurement is proposed in the paper Amplitude Information (AI) is used as a feature for state estimation of relatively low observable target (low Signal-to-Noise Ratio (SNR)) in the presence of high rate of false alarms. Rayleigh distributed noise amplitude and Swerling 3 type target model are assumed. The stochastic filter combines the Multiple Model (MM) approach with switching models for dealing with maneuvers and probabilistic association of features and measured kinematic data. The filter performance is analyzed by simulation. Results show that the suggested algorithm can track targets with SNR down to 10 dB with acceptable percentage of lost tracks, while the filter without AI works down to 13 dB. In the case of nonmaneuvering target these limits are at lower levels.

Từ khóa

#Filters #Target tracking #Artificial intelligence #Signal to noise ratio #State estimation #Noise level #Stochastic resonance #Kinematics #Performance analysis #Analytical models

Tài liệu tham khảo

van der merve, 2000, The unscented particle filter, 1 musso, 2000, Recent particle filter applied to terrain navigation, Proc of the FUSION 2000 Conference tonissen, 1998, Maximum Likelihood Track-Before-Detect with Fluctuating Target Amplitude, 33, 796 mallick, 2001, Comparison of nonlinear filtering algorithms in ground moving target indicator tracking, Proc of the FUSION 2001 Conference musick, 2001, Comparison of particle method and finite difference nonlinear filters for low SNR target tracking, Proc of the FUSION 2001 Conference kirubarajan, 1996, Low Observable Target Motion Analyzis Using Amplitude Information, 32, 1367 lerro, 1993, Interacting Multiple Model Tracking with Target Amplitude Feature, 29, 494 hue, 2001, The(MR)MTPF Particle Filters to Track Multiple Targets Using Multiple Receivers jilkov, 1996, Multiple hypothesis tracking using amplitude feature measurements, Academie des Sciences Comptes Rendus, 49, 41 difranco, 1980, Radar Detection bar-shalom, 1995, Multitarget-Multisensor Tracking Principles and Techniques bar-shalom, 1993, Estimation and Tracking Principles Techniques and Software doucet, 2000, Stochastic Sampling Algorithms for State Estimation of Jump Markov Linear Systems, 45, 188 ballantyne, 2001, A branching particle-based nonlinear filter for multi-target tracking, Proc of the FUSION 2001 Conference angelova, 2001, Application of A Monte Carlo Method for Trackimg Maneuvering Target in Clutter, 55, 15 stone, 1999, Bayesian Multiple Target Tracking helstrom, 1995, Elements of Signal Detection and Estimation doucet, 1999, On sequential Monte Carlo sampling methods for Bayesian filtering blair, 1998, Benchmark for Radar Allocation and Track Ing in ECM, 34, 1097