Improved Particle Filter for Nonlinear System State

J. Liang,X. -Y. Peng

DOI: https://doi.org/10.1049/el:20082403

2008-01-01

Electronics Letters

Abstract:A particle filter based on the similarity of the observation path is presented. When the likelihood has a bimodal nature, the particle filter leads to more accurate state estimates than SIR, APF, RPF and GPF.

What problem does this paper attempt to address?

Improved Particle Filter Based on Fine Resampling Algorithm

Bei CAO,Cai-wen MA,Zhen-tao LIU

DOI: https://doi.org/10.1016/s1005-8885(11)60253-7

2012-01-01

The Journal of China Universities of Posts and Telecommunications

Abstract:In order to solve particle degeneracy phenomenon and simultaneously avoid sample impoverishment, this paper proposed an improved particle filter based on fine resampling algorithm for general case, called as particle filter with fine resampling (PF-FR). By introducing distance-comparing process and generating new particle based on optimized combination scheme, PF-FR filter performs better than generic sampling importance resampling particle filter (PF-SIR) both in terms of effectiveness and diversity of the particle system, hence, evidently improving estimation accuracy of the state in the nonlinear/non-Gaussian models. Simulations indicate that the proposed PF-FR algorithm can maintain the diversity of particles and thus achieve the same estimation accuracy with less number of particles. Consequently, PF-FR filter is a competitive choice in the applications of nonlinear state estimation.
An Improved Particle Filtering Algorithm Using Different Correlation Coefficients for Nonlinear System State Estimation

Qingxu Meng,Kaicheng Li,Chen Zhao

DOI: https://doi.org/10.1089/big.2018.0130

IF: 4.426

Big Data

Abstract:Particle filtering (PF) algorithm has found an increasingly wide utilization in many fields at present, especially in nonlinear and non-Gaussian situations. Because of the particle degeneracy limitation, various resampling methods have been researched. This article proposed an improved PF algorithm combining with different rank correlation coefficients to overcome the shortcomings of degeneracy. By simulating iteration operation in Matlab, it discovers that the proposed algorithm provides better accuracy than sequential importance resampling, Gaussian sum particle filter, and Gaussian mixture sigma-point particle filters in Gaussian mixture noise.
The improved particle filter for object tracking

Wang Qicong,Liu Jilin

DOI: https://doi.org/10.1109/WCICA.2006.1714013

2006-01-01

Abstract:We address the problem of object tracking encountered in video processing. The proposed approach is mainly composed of object modeling, the improved particle filter and mixture filtering. First, each of visual objects can be modeled by multi-part color likelihood model. To tackle self-occlusion of the tracked objects, the color distribution representing the tracked object can be updated over time. We use the improved particle filter to ameliorate the performance of the classical particle filter. To track multiple objects simultaneously, we use multi-component mixture model whose components are modeled by the improved particle filter to form tracking algorithm Experimental results show the proposed method performs object tracking effectively. © 2006 IEEE.
Enhancing Particle Swarm Optimization Based Particle Filter Tracker

Qicong Wang,Li Xie,Jilin Liu,Zhiyu Xiang

DOI: https://doi.org/10.1007/11816171_151

2006-01-01

Abstract:A novel particle filter, enhancing particle swarm optimization based particle filter (EPSOPF), is proposed for visual tracking. Particle filter (PF) is sequential Monte Carlo simulation based on particle set representations of probability densities, which can be applied to visual tracking. However, PF has the impoverishment phenomenon which limits its application. To improve the performance of PF, particle swarm optimization with mutation operator is introduced to form new filtering, in which mutation operator maintain multiple modes of particle set and optimization-seeking procedure drives particles to their neighboring maximum of the posterior. When applied to visual tracking, the proposed approach can realize more efficient function than PF.
A Fast Estimating Linear/Nonlinear Condition Particle Filter

Aiping Huang

2009-01-01

Abstract:A fast estimating linear/non-linear condition particle filter algorithm is presented.The new method separates the related linear and non-linear conditions to overcome the RBPF(Rao-Blackwellized particle filter) algorithm shortage on processing speed in the mixed linear/non-linear model,for the estimation of non-linear condition,traditional particle filter is used,and the mean of the particles drawn from the non-linear condition is computed and propagated to the linear part to get the estimation.The simulation shows that compared with RBPF,the new method can improve the processing speed by 50%~60% and the real-time application effectively.
Particle Filter Based On Partial Accurate Observation For High Dimensional Stochastic Dynamic System

Feilong Song,Xiaohong Guan,Jiang Wu,Dong Xie,Weisheng Liu,Li Chen

DOI: https://doi.org/10.23919/ChiCC.2017.8028081

2017-01-01

Abstract:With the improvement of computer performance, particle filter as a method of state estimation, is widely used for nonlinear non-Gaussian dynamic systems. In order to reduce the partical degeneracy in the state estimation process of high-dimensional stochastic dynamic systems which commonly occurs in the conventional particle filter, an improved particle filter method based on adding more accurate direct observations of some dimensions of the system state is proposed in this paper. By arranging a plurality of sensors in the stochastic dynamic system which are able to directly and precisely measure some dimensions of the state, lots of accurate state information can be obtained. Combined the information with the inverse process of the system measurement, the particles move towards the high observation likelihood region, so that the effectiveness and accuracy of resampling process can be greatly enhenced. In the case study, a stochastic dynamic system of near-surface wind farm is established by the theory of computational fluid mechanics. Numerical results show that the proposed PF method is able to enhance the accuracy of the estimation for high dimensional stochastic dynamic system.
Efficient Particle Filter and Its Application in Contour Tracking

Peng Chen,Hui Qian,Weisong Gao,Miaoliang Zhu

DOI: https://doi.org/10.1117/12.831264

2009-01-01

Abstract:In order to improve the real-time performance of particle filter, this paper proposes an efficient particle filter algorithm and evaluates its usage in object contour tracking application. This new filter uses only one particle to predict next state in certain situations. As particle set size is one, there is no need to resample the particles before prediction. Therefore the real-time performance of particle filter is improved. To maintain the performance of particle filter, we use one particle sample only in prediction step and only in situations where particles are close enough with each other. After prediction step, the original number of particle samples is recovered. Experiment shows that the new particle filter uses less time to process the particles while retain tracking performance comparable with the regular particle filter.
Improved Particle Filter Using Clustering Similarity of the State Trajectory with Application to Nonlinear Estimation: Theory, Modeling, and Applications

Ziquan Jiao,Zhiqiang Feng,Na Lv,Wenjing Liu,Haijian Qin

DOI: https://doi.org/10.1155/2021/9916339

IF: 2.336

2021-01-01

Journal of Sensors

Abstract:A clustering similarity particle filter based on state trajectory consistency is presented for the mathematical modeling, performance estimation, and smart sensing of nonlinear systems. Starting from an information fusion model based on the consistency principle of the spatial state trajectory, the predicted observation information of the current particle filter (original trajectory) and future multistage Gaussian particle filter (modified trajectory) are selected as the state trajectories of the sampling particles. Clustering similarity methods are used to measure the state trajectories of the sampling particles and the actual system (reference trajectory). The importance weight of a first-order Markov model is updated with the measurement results. By integrating the targeted compensation scheme of the latest measurement information into the sequential importance sampling process, the adverse effects of the particle degradation phenomenon are effectively reduced. The convergence theorems of the improved particle filter are proposed and proved. The improved filter is applied to practical cases of nonlinear process estimation, economic statistical prediction, and battery health assessment, and the simulation results show that the improved particle filter is superior to traditional filters in estimation accuracy, efficiency, and robustness.
Fast Gaussian particle filtering algorithm

Peng Chen,Hui Qian,Miaoliang Zhu

2008-01-01

Abstract:A fast Gaussian particle filtering (GPF) algorithm without sampling step is presented to improve the real-time performance of GPF algorithm. The new estimator yields better performance by replacing Gaussian sampling with linear transformation. The distribution function represented by the particles of these two algorithms is identical and the computing time of linear transformation is less than that of Gaussian sampling. Particle filtering (PF) algorithm, GPF algorithm, and improved GPF algorithm are applied to two numerical examples respectively, one is a one-dimensional discrete-time nonlinear time-varying model, the other is a two-dimensional bearing only tracking (BOT) model. Simulation results are presented to demonstrate the lower complexity of the new GPF algorithm over conventional GPF algorithm and PF algorithm. Improved GPF algorithm can generate 1000 particles in 22 ms, 6 ms less than that of GPF algorithm with the same estimation accuracy.
Improved design of proposal distribution for particle filter

LIU Zhen-tao,Bei Cao,SHEN Xu-bang,刘镇弢,曹蓓,沈绪榜

2012-01-01

Abstract:One crucial issue in particle filtering is the selection of proposal distribution. Good proposal can effectively alleviate particle degeneracy and thus improve filtering accuracy. In this paper, we propose a new type of proposal distribution for particle filter, called as R-IEKF proposal. By combining iterated extended kalman filter with Rauch-Tung-Striebel optimal smoother, the new proposal integrates the latest observation into system and approximates the true posterior distribution reasonably well, hence generating more precise and stable particles against measurement noise. The simulation results indicate that the improved particle filter with R-IEKF proposal prevails over PF-EKF and UPF both in tracking accuracy and filtering stability. Consequently, PF-RIEKF is a competitive choice in noisy measurement environment.
Estimating States for Nonlinear Systems Using the Particle Filter

Johnny Condori,Amin Maghareh,Shirley Dyke

DOI: https://doi.org/10.48550/arXiv.1911.03263

2019-11-08

Abstract:Kalman filtering has been traditionally applied in three application areas of estimation, state estimation, parameter estimation (a.k.a. model updating), and dual estimation. However, Kalman filter is often not sufficient when experimenting with highly uncertain nonlinear dynamic systems. In this study, a nonlinear estimator is developed by adopting a particle filter algorithm that takes advantage of measured signals. This approach is shown to significantly improve the ability to estimate states. To illustrate this approach, a model for a nonlinear device coupled with a hydraulic actuator plays the role of an actual plant and a nonlinear algebraic function is considered as an approximation of the nonlinear device, thus generating non-parametric and parametric uncertainties. Then we use displacement and force signals to improve the distribution of the states by resampling the set of particles. Finally, all of the states are estimated from these posterior density functions. A set of simulations considering three different noise levels demonstrates that the performance of the particle filter approach is superior to the Kalman filter, yielding substantially better performance when estimating nonlinear physical systems in the presence of modeling uncertainties.

Systems and Control
Particle Filter with Fine Resampling for Bearings-Only Tracking

Bei Cao,Caiwen Ma,Zhentao Liu

DOI: https://doi.org/10.1016/j.proeng.2012.01.553

2012-01-01

Procedia Engineering

Abstract:To solve the nonlinear filtering problem in the Bearings-Only Tracking (BOT), an improved particle filter with fine resampling is proposed, called as PF-FR (Particle Filter with Fine Resampling). By introducing distance-comparing process and generating new particle based on optimized combination scheme, PF-FR filter performs better than generic PF-SIR filter both in terms of effectiveness and diversity of the particle system, hence, evidently enhancing tracking accuracy in bearings-only tracking problem. Simulations indicate that the proposed PF-FR algorithm is effective.
A new kind of particle filter algorithm based on immune particle swarm optimization in objective tracking

Longhua Ma,Yu Zhang,Hui Li,Zheming Lu

2013-01-01

Abstract:Nowadays researchers always adopt extended Kalman filter (EKF) and unscented Kalman filter (UKF) to deal with the objective tracking problem in the nonlinear situation, while in reality, system is often affected by the flicker noise which will lead to low accuracy and instability of filters. Though particle filter can deal with the problems of nonlinear and flicker noise, it cannot avoid the problems of particle degradation and sample dilution. Concerning the problems above in particle filter, a new kind of particle filter algorithm based on immune particle swarm optimization was presented and we introduced it into the field of two-dimensional radar objective tracking. The simulation results prove the effectiveness of the proposed algorithm by reducing the tracking error, improving filtering accuracy and avoiding problems of sample dilution. © 2013 ICIC International.
An integrated weights particle filter algorithm based on correlation particle estimation and sequential importance re-sampling

Tao Zhang,Daixi Shi,Zhiyong Xie,Jingyan Song

DOI: https://doi.org/10.1109/CISP.2013.6743852

2013-01-01

Abstract:In this paper, a new integrated weight particle filter (IWPF) algorithm is proposed based on the combination of correlation particle estimation (CPE) weight and sequential importance re-sampling (SIR) weight. This method can reduce degeneracy phenomenon and re-sampling times of traditional particle filter. By choosing the typical nonlinear system model, the simulation results show that IWPF performs better than CPE and SIR. In our simulation case, this method can provide a 15% increase of accuracy in state estimation and a 30% decrease of re-sampling times.
Particle Filter with Hybrid Proposal Distribution for Nonlinear State Estimation.

Fasheng Wang,Yuejin Lin,Tao Zhang,Jingbo Liu

DOI: https://doi.org/10.4304/jcp.6.11.2491-2501

2011-01-01

Abstract:Particle filters have been widely used in solving nonlinear filtering problems. Proposal Distribution design is a key issue for these methods and has vital effect on simulation results. Various proposal distributions have been proposed to improve the performance of particle filters, but practical situations have promoted the researchers to design better candidate for proposal distributions in order to gain better performance. This paper proposes a hybrid proposal distribution designed for particle filtering framework. The algorithm uses hybrid Kalman filter to generate the proposal distribution, which make efficient use of the latest observations and generate more closed approximation of the posterior probability density. The yielded algorithm is named as hybrid Kalman particle filter. In the experiments, we use a scalar estimation model and a real world application problem to evaluate the new algorithm. The experimental results show that the new algorithm outperforms other algorithms with different proposal distributions. In order to reduce the time consumption of the new algorithm, we propose an improvement strategy, namely partition-conquer strategy, in which the particles are partitioned into two parts, with one part drawn from the hybrid Kalman filter, another part from the transition prior. The validity of the strategy is verified through an experiment.
A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking

M. Sanjeev Arulampalam,Simon Maskell,Neil Gordon,Tim Clapp,M.S. Arulampalam,S. Maskell,N. Gordon,T. Clapp

DOI: https://doi.org/10.1109/78.978374

IF: 4.875

2002-01-01

IEEE Transactions on Signal Processing

Abstract:Increasingly, for many application areas, it is becoming important to include elements of nonlinearity and non-Gaussianity in order to model accurately the underlying dynamics of a physical system. Moreover, it is typically crucial to process data on-line as it arrives, both from the point of view of storage costs as well as for rapid adaptation to changing signal characteristics. In this paper, we review both optimal and suboptimal Bayesian algorithms for nonlinear/non-Gaussian tracking problems, with a focus on particle filters. Particle filters are sequential Monte Carlo methods based on point mass (or “particle”) representations of probability densities, which can be applied to any state-space model and which generalize the traditional Kalman filtering methods. Several variants of the particle filter such as SIR, ASIR, and RPF are introduced within a generic framework of the sequential importance sampling (SIS) algorithm. These are discussed and compared with the standard EKF through an illustrative example.

engineering, electrical & electronic
An Improved Particle Filter For Target Tracking In Sensor Systems

Xue Wang,Sheng Wang,Jun-Jie Ma

DOI: https://doi.org/10.3390/s7010144

IF: 3.9

2007-01-01

Sensors

Abstract:Sensor systems are not always equipped with the ability to track targets. Sudden maneuvers of a target can have a great impact on the sensor system, which will increase the miss rate and rate of false target detection. The use of the generic particle filter (PF) algorithm is well known for target tracking, but it can not overcome the degeneracy of particles and cumulation of estimation errors. In this paper, we propose an improved PF algorithm called PF-RBF. This algorithm uses the radial-basis function network (RBFN) in the sampling step for dynamically constructing the process model from observations and updating the value of each particle. With the RBFN sampling step, PF-RBF can give an accurate proposal distribution and maintain the convergence of a sensor system. Simulation results verify that PF-RBF performs better than the Unscented Kalman Filter (UKF), PF and Unscented Particle Filter (UPF) in both robustness and accuracy whether the observation model used for the sensor system is linear or nonlinear. Moreover, the intrinsic property of PF-RBF determines that, when the particle number exceeds a certain amount, the execution time of PF-RBF is less than UPF. This makes PF-RBF a better candidate for the sensor systems which need many particles for target tracking.
An Improved Gaussian Mixture Particle Filter Based Targets Tracking Algorithm for Bearing-Only Tracking System

Zhentao Liu

2012-01-01

Abstract:An improved Gaussian mixture particle filter algorithm is proposed for the highly non-linear bearing-only tracking system where the common tracking filters often fail to catch and keep tracking of the emitter.In the algorithm,based on the characteristics of SPKF and particle filter,the limited Gaussian mixture model is used to approximate the posterior density of states,system noise and measurement noise.The greedy EM is used to obtain the reduced order model and overcome such disadvantages of the standard EM as the number of the mixture components is assumed a known a priori,the performance of the overall parameter estimation process depends on the given good initial settings,and the estimated parameter can be resulted from some local optimum points,thus lessening effects caused by sampling depletion.Simulation results show that the algorithm outperforms the one based on PF and the one based on EM-GMPF in tracking accuracy and stability.Therefore it is more suitable to the nonlinear state estimation.
A Fine Resampling Algorithm for General Particle Filters

Bei Cao,CaiWen Ma,ZhenTao Liu

DOI: https://doi.org/10.1109/cisp.2011.6100779

2011-01-01

Abstract:Resampling algorithm for particle filters aimed at solving particle degeneracy problem but causing sample impoverishment. To maintain the diversity of the particle system and thus enhance the accuracy of state estimation in the nonlinear/non-Gaussian models, the paper proposed an improved resampling algorithm for general particle filters, called as PF-FR (Particle Filters with Fine Resampling). By introducing distance-comparing process and generating new particle based on optimized combination scheme, PF-FR algorithm overcomes the defects of general resampling algorithm. Simulations prove that the proposed PF-FR algorithm can effectively avoid sample impoverishment and simultaneously solve the degeneracy problem, thereby improving the estimation accuracy evidently. As a consequence, PF-FR algorithm is an effective alternative to resampling algorithm.
A Novel Particle Filter Aided By Interval Analysis Approach

Yang Dong-Fang,Sun Fu-Chun,Wang Shi-Cheng,Min Hai-Bo

DOI: https://doi.org/10.1080/03610918.2012.683922

2013-01-01

Communications in Statistics - Simulation and Computation

Abstract:This paper addresses the issue of state estimation in nonlinear systems in the presence of non-Gaussian and bounded noises, under which the interval analysis based estimation is introduced as an auxiliary approach of generic particle filter (PF). This yields the so-called Set-Membership aided particle filter (SMPF). Unlike the mature alternatives of generic particle filter, the proposal distribution of SMPF approximates the posterior probability density function (PDF), not only on the numerical value but also on the definition-domain, and the performance analysis on the proposed alternative is proven through detailed formulations. In addition, contrasting simulations under SMPF and other mature alternatives also validate the effectiveness of SMPF.

Improved Particle Filter for Nonlinear System State

Improved Particle Filter Based on Fine Resampling Algorithm

An Improved Particle Filtering Algorithm Using Different Correlation Coefficients for Nonlinear System State Estimation

The improved particle filter for object tracking

Enhancing Particle Swarm Optimization Based Particle Filter Tracker

A Fast Estimating Linear/Nonlinear Condition Particle Filter

Particle Filter Based On Partial Accurate Observation For High Dimensional Stochastic Dynamic System

Efficient Particle Filter and Its Application in Contour Tracking

Improved Particle Filter Using Clustering Similarity of the State Trajectory with Application to Nonlinear Estimation: Theory, Modeling, and Applications

Fast Gaussian particle filtering algorithm

Improved design of proposal distribution for particle filter

Estimating States for Nonlinear Systems Using the Particle Filter

Particle Filter with Fine Resampling for Bearings-Only Tracking

A new kind of particle filter algorithm based on immune particle swarm optimization in objective tracking

An integrated weights particle filter algorithm based on correlation particle estimation and sequential importance re-sampling

Particle Filter with Hybrid Proposal Distribution for Nonlinear State Estimation.

A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking

An Improved Particle Filter For Target Tracking In Sensor Systems

An Improved Gaussian Mixture Particle Filter Based Targets Tracking Algorithm for Bearing-Only Tracking System

A Fine Resampling Algorithm for General Particle Filters

A Novel Particle Filter Aided By Interval Analysis Approach