Soft systematic resampling for accurate posterior approximation and increased information retention in particle filtering

A fast-track smoothing algorithm based on the KD Tree is proposed to address the problem of state estimation in underwater multi-target tracking. For multi-target tracking, the multi-Bernoulli filter becomes intractable due to the high-dimensional nature of the posterior. This problem can be overcome by the particle filter (PF), which approximates the posterior by a set of weighted samples (“particles”). The probability distribution of the multi-Bernoulli algorithm based on this algorithm contains a series of particles with weights. Therefore, while smoothing it, the state transfer probability of each particle must be computed, and the state transfer function must be calculated, which is related to the difference in states between the k-1 moment and the k moment, we refer to this difference as the state distance. However, the KD Tree can be used to divide the particle set into different state spaces according to the particle states, and thus calculate the state distances more efficiently. The remarkable feature of this algorithm is that the time complexity is greatly reduced from that O(N2) of the traditional algorithm to O(NlogN). This paper implements the proposed method using particles, the improved algorithm based on KD tree effectively mitigates the disadvantages of previous forward-backward smoothing algorithms, and significantly reduces the computation time while improving the system state estimation precision. Experimental results of simulation analysis and the optimal sub-patten assignment distance is employed as an assessment criterion to validate the algorithm’s performance.

show abstract

“…, and the ( 12) is substituted into the (7). ( 15) to (18) [16] illustrate how to derive the labeled multi-Bernoulli smoothing equation.…”

Section: Principle Of Lmb Algorithmmentioning

confidence: 99%

Fast Track Smoothing Algorithm for Underwater Target Tracking Based on KD Tree

Jin

Wang

et al. 2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…The soft systematic resampling approach [278] presented in this section is an extension and correction to the soft resampling technique and presents a scheme to overcome the problem of discarding the low weight particles all the time. The main idea is to first use a modified variant of soft resampling without discarding any particles and then stochastically resample the lower weights using the systematic resampler [52,136].…”

Section: Motivation For Soft Systematic Resamplingmentioning

confidence: 99%

“…The systematic resampler is used to stochastically resample a few lower weight particles so that not every low weight particle is eliminated all the time. The consequential soft systematic resampling [278] in presented here.…”

Section: Soft Systematic Resampling Algorithmmentioning

confidence: 99%

Bayesian multiple target tracking

Choppala¹

View full text Add to dashboard Cite

<p>This thesis addresses several challenges in Bayesian target tracking, particularly for array signal processing applications, and for multiple targets. The optimal method for multiple target tracking is the Bayes’ joint filter that operates by hypothesising all the targets collectively using a joint state. As a consequence, the computational complexity of the filter increases rapidly with the number of targets. The probability hypothesis density and the multi-Bernoulli filters that overcome this complexity do not possess a suitable framework to operate directly on phased sensor array data. Instead, such data is converted into beamformer images in which close targets may not be effectively resolved and much information is lost. This thesis develops a multiple signal classification (MUSIC) based multi-target particle filter that improves upon the filters mentioned above. A MUSIC based multi-Bernoulli particle filter is also developed, that operates more directly on array data. The above mentioned particle filters require a resampling step which impedes information accumulation over successive observations, and affects the detection of very covert targets. This thesis develops soft resampling and soft systematic resampling to overcome this problem without affecting the accuracy of approximation. Additionally, modified Kolmogorov-Smirnov testing is proposed, to numerically evaluate the accuracy of the particle filter approximation.</p>

show abstract

Auxiliary particle filtering with lookahead support for univariate state space models

Choppala

2024

Ann Univ Ferrara

View full text Add to dashboard Cite

Soft systematic resampling for accurate posterior approximation and increased information retention in particle filtering

Cited by 7 publications

References 16 publications

Fast Track Smoothing Algorithm for Underwater Target Tracking Based on KD Tree

Fast Track Smoothing Algorithm for Underwater Target Tracking Based on KD Tree

Bayesian multiple target tracking

Auxiliary particle filtering with lookahead support for univariate state space models

Contact Info

Product

Resources

About