Conventional Particle Filter Research Articles

A joint particle filter and expectation maximization approach to machine condition prognosis

This paper presents a probabilistic model based approach for machinery condition prognosis based on particle filter by integrating physical knowledge with in-process measurements into a state space framework to account for uncertainty and nonlinearity in machinery degradation process. One limitation of conventional particle filter is that condition prognosis is performed based on the model with predetermined parameters obtained from simulation studies or lab-controlled tests. Due to the stochastic nature of machinery defect propagation under varying operating conditions, model parameters may vary in practice which causes prediction errors. To address it, an integrated state prediction and parameter estimation framework based on particle filter and expectation-maximization algorithm is formulated and investigated. The model parameters are adaptively estimated based on expectation-maximization algorithm utilizing hidden degradation state and available in-process measurements. Particle filter is then performed on the identified model with estimated parameters following Bayesian inference scheme to improve the robustness and accuracy of machinery condition prognosis. The effectiveness of the developed method is demonstrated through a simulation study and an experimental run-to-failure bearing test in a wind turbine.

Utilizing Time Redundancy for Particle Filter-Based Transfer Alignment

Signal detection in the presence of high noise is a challenge in natural sciences. From understanding signals emanating out of deep space probes to signals in protein interactions for systems biology, domain specific innovations are needed. The present work is in the domain of transfer alignment (TA), which deals with estimation of the misalignment of deliverable daughter munitions with respect to that of the delivering mother platform. In this domain, the design of noise filtering scheme has to consider a time varying and nonlinear system dynamics at play. The accuracy of conventional particle filter formulation suffers due to deviations from modeled system dynamics. An evolutionary particle filter can overcome this problem by evolving multiple system models through few support points per particle. However, this variant has even higher time complexity for real-time execution. As a result, measurement update gets deferred and the estimation accuracy is compromised. By running these filter algorithms on multiple processors, the execution time can be reduced, to allow frequent measurement updates. Such scheme ensures better system identification so that performance improves in case of simultaneous ejection of multiple daughters and also results in better convergence of TA algorithms for single daughter.

Robust tracking of multiple persons in real-time video

In this paper, we present a robust person tracking method that the particle swarm optimization (PSO) algorithm is used as the tracking strategy. The method is divided into two procedures: object/background segmentation and tracking. For object/background segmentation, we use the temporal differencing to detect the regions of interest. For tracking, the PSO algorithm is used for overcome the robustness problem in the high noisy background and multiple moving persons and/or under occlusion. The particles in PSO represent the position, width and height of the search window, and the fitness function is calculated by the distance of the color feature vector and the histogram intersection. When occluded, we add the motion vector plus the previous position of the tracking model. The particles fly around the search region to obtain an optimal match of the target. The experiments show that the proposed method can track the single person, multiple people even when occluded, and is more efficient and accurate than the conventional particle filter method.

Oscillations-free PID control of anesthetic drug delivery in neuromuscular blockade

Background and ObjectivesThe PID-control of drug delivery or the neuromuscular blockade (NMB) in closed-loop anesthesia is considered. The NMB system dynamics portrayed by a Wiener model can exhibit sustained nonlinear oscillations under realistic PID gains and for physiologically feasible values of the model parameters. Such oscillations, also repeatedly observed in clinical trials, lead to under- and over-dosing of the administered drug and undermine patient safety. This paper proposes a tuning policy for the proportional PID gain that via bifurcation analysis ensures oscillations-free performance of the control loop. Online estimates of the Wiener model parameters are needed for the controller implementation and monitoring of the closed-loop proximity to oscillation. MethodsThe nonlinear dynamics of the PID-controlled NMB system are studied by bifurcation analysis. A database of patient models estimated under PID-controlled neuromuscular blockade during general anesthesia is utilized, along with the corresponding clinical measurements. The performance of three recursive algorithms is compared in the application at hand: an extended Kalman filter, a conventional particle filter (PF), and a PF making use of an orthonormal basis to estimate the probability density function from the particle set. ResultsIt is shown that with a time-varying proportional PID gain, the type of equilibria of the closed-loop system remains the same as in the case of constant controller gains. The recovery time and frequency of oscillations are also evaluated in simulation over the database of patient models. Nonlinear identification techniques based on model linearization yield biased parameter estimates and thus introduce superfluous uncertainty. The bias and variance of the estimated models are related to the computational complexity of the identification algorithms, highlighting the superiority of the PFs in this safety-critical application. ConclusionsThe study demonstrates feasibility of the proposed oscillation-free control strategy combining bifurcation theory based design and online parameter estimation by PF.

Enhanced particle filter for states and parameters estimation in structural health monitoring applications

In this paper, an iterated square-root central difference Kalman particle filter method (ISRCDKF-PF) is used for the estimation of the state variables and model parameters of nonlinear structural systems. In the current work, we propose to extend our previous work (Mansouri et al. in J Civil Struct Health Monit 5(4):493–508, 2015) to deal with non-parametric Monte Carlo sampling-based method and propose to use an enhanced PF technique which incorporates the latest observations into a prior updating scheme using the ISRCDKF algorithm. Various conventional and state-of-the-art state estimation methods are compared for the estimation performance, namely the unscented Kalman filter (UKF), the square-root central difference Kalman filter (SRCDKF), the iterated unscented Kalman filter (IUKF), the iterated square-root central difference Kalman filter (SRCDKF), the conventional particle filter (PF), the unscented Kalman particle filter (UKF-PF), the SRCDKF-PF, the iterated unscented Kalman particle filter (IUKF-PF) and the developed ISRCDKF-PF, in two comparative studies through two examples, one using synthetic data and the other using simulated three DOF damped system data. In the first comparative study, the state variables are estimated from noisy measurements of these variables, and the comparison of the different estimation techniques is performed by computing the root mean square error (RMSE) of the state with respect to the noise-free data. In the second comparative study, both the state variables and the model parameters are simultaneously estimated, and the impact of the used measurement noise, and number of estimated states/parameters on the performances of the estimation techniques are investigated. The ISRCDKF-PF algorithm consists of a PF based on ISRCDKF to obtain a better importance proposal distribution. This proposal is able to integrate the latest observation into the state density, then it can improve the posteriori density. The results of both comparative study show that PF, UKF-PF, SRCDKF-PF, IUKF-PF and ISRCDKF-PF provide improved estimation performance over the UKF, SRCDKF, IUKF, ISRCDKF. The results also show that ISRCDKF-PF provides improved estimation performance over IUKF-PF, even with abrupt changes in estimated states, and both of them provide better accuracy than the conventional PF, UKF-PF and SRCDKF-PF. These advantages of the ISRCDKF-PF are due to the fact that it uses an optimal proposal distribution which make efficient use of the latest observation by using the ISRCDKF algorithm.

An Ensemble Kushner-Stratonovich-Poisson Filter for Recursive Estimation in Nonlinear Dynamical Systems

We propose a Monte Carlo filter for recursive estimation of diffusive processes that modulate the instantaneous rates of Poisson measurements. A key aspect is the additive update, through a gain-like correction term, empirically approximated from the innovation integral in the time-discretized Kushner-Stratonovich equation. The additive filter-update scheme eliminates the problem of particle collapse encountered in many conventional particle filters. Through a few numerical demonstrations, the versatility of the proposed filter is brought forth.

Maneuvering target track-before-detect via multiple-model Bernoulli particle filter

Target tracking using non-threshold raw data with low signal-to-noise ratio is a very difficult task, and the model uncertainty introduced by target’s maneuver makes it even more challenging. In this work, a multiple-model based method was proposed to tackle such issues. The method was developed in the framework of Bernoulli filter by integrating the model probability parameter and implemented via sequential Monte Carlo (particle) technique. Target detection was accomplished through the estimation of target’s existence probability, and the estimate of target state was obtained by combining the outputs of modeldependent filtering. The simulation results show that the proposed method performs better than the TBD method implemented by the conventional multiple-model particle filter.

Mining Evolving Data Streams with Particle Filters

We propose a particle filter‐based learning method, PF‐LR, for learning logistic regression models from evolving data streams. The method inherently handles concept drifts in a data stream and is able to learn an ensemble of logistic regression models with particle filtering. A key feature of PF‐LR is that in its resampling, step particles are sampled from the ones that maximize the classification accuracy on the current data batch. Our experiments show that PF‐LR gives good performance, even with relatively small batch sizes. It reacts to concept drifts quicker than conventional particle filters while being robust to noise. In addition, PF‐LR learns more accurate models and is more computationally efficient than the gradient descent method for learning logistic regression models. Furthermore, we evaluate PF‐LR on both synthetic and real data sets and find that PF‐LR outperforms some other state‐of‐the‐art streaming mining algorithms on most of the data sets tested.

Robust visual tracking based on joint multi-feature histogram by integrating particle filter and mean shift

In this paper, we proposed a robust visual tracking and simultaneous recognition method in a particle filter frame work. First, an appearance model based on Multi-Part Joint Color Texture and Edge of Orientation Histogram with corrected background weighted histogram is Build, which is adopted to calculate the likelihood of particles; then, perform scale invariant feature transform (SIFT) keypoint detection and matching to predict the coarse position of the target; thirdly, fine tune the target position by enlarged searching region to improve the proposal distribution of particle filter, mean shift is embedded into particle filters, in which a smaller number of samples is used to estimate the posterior distribution than conventional particle filters by shifting samples to their neighboring modes of the observation so that samples are moved to have large weights. Finally, estimating the target position. Experimental results demonstrate that this algorithm can track the object accurately in conditions of scale modifications, rotation, abrupt shifts, as well as clutter and partial occlusions occurring to the tracking object with good robustness.

Pose estimation for planar contact manipulation with manifold particle filters

We investigate the problem of using contact sensors to estimate the pose of an object during planar pushing by a fixed-shape hand. Contact sensors are unique because they inherently discriminate between “contact” and “no-contact” configurations. As a result, the set of object configurations that activates a sensor constitutes a lower-dimensional contact manifold in the configuration space of the object. This causes conventional state estimation methods, such as the particle filter, to perform poorly during periods of contact due to particle starvation. In this paper, we introduce the manifold particle filter as a principled way of solving the state estimation problem when the state moves between multiple manifolds of different dimensionality. The manifold particle filter avoids particle starvation during contact by adaptively sampling particles that reside on the contact manifold from the dual proposal distribution. We describe three techniques, one analytical and two sample-based, of sampling from the dual proposal distribution and compare their relative strengths and weaknesses. We present simulation results that show that all three techniques outperform the conventional particle filter in both speed and accuracy. In addition, we implement the manifold particle filter on a real robot and show that it successfully tracks the pose of a pushed object using commercially available tactile sensors.

Particle filter algorithms for identification of minimally parametrized Wiener models of drug administration effect

Closed-loop drug administration is performed by a feedback controller keeping the patient response around a pre-set level in the face of process disturbance and uncertainty. To guarantee the safe operating envelope of the closed-loop drug delivery system, controller design for performance and robustness is typically based on a mathematical model of the plant. Minimally parametrized Wiener models comprising a linear dynamic block and a smooth output nonlinearity have been shown to accurately capture the patient-specific pharmacokinetics and pharmacodynamics in closed-loop drug administration. Due to the nonlinear model dynamics, identification techniques based on linear or linearized models yield biased parameter estimates and thus introduce superfluous uncertainty. The present paper compares the performance of three nonlinear system identification algorithms: an extended Kalman filter, a conventional particle filter (PF), and a PF making use of a orthonormal basis to estimate the probability density function from the particle set. A database of patient models estimated under PID-controlled neuromuscular blockade during general anesthesia is utilized, along with the clinical measurements. The bias and variance of the estimated models are related to the computational complexity of the identification algorithms, highlighting the superiority of the PFs in this safety-critical application.

Lane detection and tracking based on annealed particle filter

This paper describes a highway lane detection and tracking algorithm based on an annealed particle filter; the algorithm combines multiple cues of road images with the annealed particle filter. We build a geometric lane model that can be applied to not only linear roads but also to curved roads. As a first step, preprocessing with a bar filter and color cues is used. In the annealed particle filter step, a K-means algorithm is utilized to measure the weights of the particles. We realize lane detection and tracking using the annealed particle filter which is the main contribution of the current paper. Experimental results show that our method is effective in various road situations. What’s more, the particle number and time cost of the annealed particle filter for each frame is largely reduced compared with those values when using the conventional particle filter.

An Improved Particle Filter SLAM Algorithm in Similar Environments

Because of the challenging data association in similar environments, a large number of particles are needed to improve the precision in particle filtering SLAM (simultaneous localization and mapping).An improved particle filter SLAM algorithm based on particle swarm optimization in similar environments is proposed. A multimode proposal distribution is acquired by combining the information of the odometry and the laser scanning. Particles are concentrated to the region of each posterior probability distribution maximum value by PSO. The performance of the conventional particle filter SLAM is improved. The simulation experiment results prove its effectiveness and feasibility.

PARTICLE FILTER BASED VEHICLE TRACKING APPROACH WITH IMPROVED RESAMPLING STAGE

ICTACT Journals are peer reviewed International Journals published quarterly. With the objective of creating Indian-based Journals of International quality and to build up local research content, ICTACT Journals exists for the purpose of providing a quality peer-reviewed platform for Scientists, Researchers, Academicians and Engineers who carry out their original research. The Journals will highlight the continued growth and new challenges in both basic research and application development.

GMSPPF for Magnetic Ellipsoid Tracking with a Magnetic Gradiometer Tensor

Magnetic ellipsoid tracking problem is characterized by high nonlinearity. In this study, the determination of target position, magnetic moment, and velocity is formulated as a Bayesian estimation problem for dynamic systems, a recursive approach is proposed to estimate the trajectory and magnetic moment component of the target using data collected with a magnetic gradiometer tensor. Particle filter provides a solution to this problem. In addition to the conventional particle filter, the proposed tracking and classification algorithm uses Gaussian mixed mode to represent the posterior state density of the unknown parameters, which is named as Gaussian mixture sigma point particle filters(GMSPPF). The performance of the proposed method has been evaluated through simulation experiment. The results indicate that the method has achieved the magnetic ellipsoid tracking and GMSPPF has better estimation performance and less computational complexity than other related algorithms.

Visual Multi-Object Tracking in the Presence of Cluttered Scenes

The aim of this study was to investigate the visual multi-object tracking in the presence of cluttered scenes. A improved algorithm of fusing multi-source information including location and color evidences were introduced based on Dezert-Smarandache Theory (DSmT) and Particle Filters (PF). Results showed that the conflict strategy and DSmT combination model were available and the proposed approach exhibited a significantly better performance for dealing with high conflict between evidences than a conventional PF. The suggested approach can easily be generalized to deal with larger number of visual multi-object and additional cues in the presence of cluttered scenes.

Annealed Particle Filter Algorithm Used for Lane Detection and Tracking

This paper describes a lane detection and tracking method based on annealed particle filter algorithm, which combines multiple cues with annealed particle filter. As a first step, preprocessing, with bar filter and color cues being used. In the annealed particle filter step, angle information of edge map is utilized to measure weights of particles. Experiments show that the time cost of annealed particle filter algorithm for each frame is largely reduced comparing with the lane detection and tracking using conventional particle filter algorithm, which is the main contribution of this paper. Furthermore, on this basis, we build a robust lane model which can be applied to not only the linear road but also the curved road. The experiments indicate that it is effective for lane detection and tracking. 

Object Tracking with an Evolutionary Particle Filter Based on Self-Adaptive Multi-Features Fusion

Particle filter algorithms are widely used for object tracking in video sequences, but the standard particle filter algorithm cannot solve the validity of particles ideally. To solve the problems of particle degeneration and sample impoverishment in a particle filter tracking algorithm, an improved object tracking algorithm is proposed, which combines a multi-feature fusion method and a genetic evolution mechanism. The algorithm dynamically computes the feature's fusion weight by the discriminability of each vision feature and then constructs the important density function based on selecting a feature's fusion method adaptively. Moreover, a self-adaptive genetic evolutionary mechanism is introduced into the particle resampling process and makes the particle become an agent with the ability of dynamic self-adaption. With self-adaptive crossover and mutation operators, the evolution system produces a large number of new particles, which can better approximate the true state of the tracking object. The experimental results show that the proposed object tracking algorithm surpasses the conventional particle filter on both robustness and accuracy, even though the tracking object is very challenging regarding illumination variation, structural deformation, the interference of similar targets and occlusion.

Particle Filter with Gaussian Weighting for Human Tracking

Particle filter for object tracking could achieve high tracking accuracy. To track the object, this method generates a number of particles which is the representation of the candidate target object. The location of target object is determined by particles and each weight. The disadvantage of conventional particle filter is the computational time especially on the computation of particle’s weight. Particle filter with Gaussian weighting is proposed to accomplish the computational problem. There are two main stages in this method, i.e. prediction and update. The difference between the conventional particle filter and particle filter with Gaussian weighting is in the update Stage. In the conventional particle filter method, the weight is calculated in each particle, meanwhile in the proposed method, only certain particle’s weight is calculated, and the remain particle’s weight is calculated using the Gaussian weighting. Experiment is done using artificial dataset. The average accuracy is 80,862%. The high accuracy that is achieved by this method could use for the real-time system tracking

Efficient particle filter algorithm for ultrasonic sensor-based 2D range-only simultaneous localisation and mapping application

Owing to low cost and relatively accurate range measurement, ultrasonic sensors are widely used in various simultaneous localisation and mapping (SLAM) applications. In spite of the abundance of ultrasonic sensor based SLAM applications, a simple and efficient algorithm for an ultrasonic sensor based positioning system with good accuracy and low computational complexity has not yet emerged. The major difficulty is the trade-off between localisation accuracy and computational complexity in most SLAM algorithms, such as extended Kalman filter (EKF) and particle filter. Typically, they improve localisation accuracy by increasing the density of the landmarks, as a result leading to high computational complexity of algorithms and limiting the utilisation of algorithms into online SLAM systems. This study addresses an improved particle filter algorithm to solve ultrasonic sensor based 2D range-only SLAM problem with relatively good accuracy and low computational complexity. This algorithm uses a simple four fixed features based system model to limit the density of the landmarks. A technique called map adjustment is proposed to increase the accuracy and efficiency of the algorithm. Using map adjustment, the proposed particle filter algorithm can improve localisation accuracy and lower computational complexity. The experiment results demonstrate that this algorithm has a better performance than conventional particle filter localisation algorithm.