Showing papers on "Particle filter published in 2019"

Learning Multi-Task Correlation Particle Filters for Visual Tracking

Abstract: In this paper, we propose a multi-task correlation particle filter (MCPF) for robust visual tracking. We first present the multi-task correlation filter (MCF) that takes the interdependencies among different object parts and features into account to learn the correlation filters jointly. Next, the proposed MCPF is introduced to exploit and complement the strength of a MCF and a particle filter. Compared with existing tracking methods based on correlation filters and particle filters, the proposed MCPF enjoys several merits. First, it exploits the interdependencies among different features to derive the correlation filters jointly, and makes the learned filters complement and enhance each other to obtain consistent responses. Second, it handles partial occlusion via a part-based representation, and exploits the intrinsic relationship among local parts via spatial constraints to preserve object structure and learn the correlation filters jointly. Third, it effectively handles large scale variation via a sampling scheme by drawing particles at different scales for target object state estimation. Fourth, it shepherds the sampled particles toward the modes of the target state distribution via the MCF, and effectively covers object states well using fewer particles than conventional particle filters, thereby resulting in robust tracking performance and low computational cost. Extensive experimental results on four challenging benchmark datasets demonstrate that the proposed MCPF tracking algorithm performs favorably against the state-of-the-art methods.

Particle filters for high-dimensional geoscience applications: A review

Abstract: Particle filters contain the promise of fully nonlinear data assimilation. They have been applied in numerous science areas, including the geosciences, but their application to high-dimensional geoscience systems has been limited due to their inefficiency in high-dimensional systems in standard settings. However, huge progress has been made, and this limitation is disappearing fast due to recent developments in proposal densities, the use of ideas from (optimal) transportation, the use of localization and intelligent adaptive resampling strategies. Furthermore, powerful hybrids between particle filters and ensemble Kalman filters and variational methods have been developed. We present a state-of-the-art discussion of present efforts of developing particle filters for high-dimensional nonlinear geoscience state-estimation problems, with an emphasis on atmospheric and oceanic applications, including many new ideas, derivations and unifications, highlighting hidden connections, including pseudo-code, and generating a valuable tool and guide for the community. Initial experiments show that particle filters can be competitive with present-day methods for numerical weather prediction, suggesting that they will become mainstream soon.

An Improved Particle Filter With a Novel Hybrid Proposal Distribution for Quantitative Analysis of Gold Immunochromatographic Strips

Abstract: In this paper, a novel statistical pattern recognition method is proposed for accurately segmenting test and control lines from the gold immunochromatographic strip (GICS) images for the benefits of quantitative analysis. A new dynamic state-space model is established, based on which the segmentation task of test and control lines is transformed into a state estimation problem. Especially, the transition equation is utilized to describe the relationship between contour points on the upper and the lower boundaries of test and control lines, and a new observation equation is developed by combining the contrast of between-class variance and the uniformity measure. Then, an innovative particle filter (PF) with a hybrid proposal distribution, namely, deep-belief-network-based particle filter (DBN-PF) is put forward, where the deep belief network (DBN) provides an initial recognition result in the hybrid proposal distribution, and the particle swarm optimization algorithm moves particles to regions of high likelihood. The performance of proposed DBN-PF method is comprehensively evaluated on not only an artificial dataset but also the GICS images in terms of several indices as compared to the PF and DBN methods. It is demonstrated via experiment results that the proposed approach is effective in quantitative analysis of GICS.

Fault diagnosis of gearbox based on RBF-PF and particle swarm optimization wavelet neural network

Abstract: The gear cracks of gear box are one of most common failure forms affecting gear shaft drive. It has become significant for practice and economy to diagnose the situation of gearbox rapidly and accurately. The extracted signal is filtered first to eliminate noise, which is pretreated for the diagnostic classification based on the particle filter of radial basis function. As traditional error back-propagation of wavelet neural network with falling into local minimum easily, slow convergence speed and other shortcomings, the particle swarm optimization algorithm is proposed in this paper. This particle swarm algorithm that optimizes the weight values of wavelet neural network (scale factor) and threshold value (the translation factor) was developed to reduce the iteration times and improve the convergence precision and rapidity so that the various parameters of wavelet neural network can be chosen adaptively. Experimental results demonstrate that the proposed method can accurately and quickly identify the damage situation of the gear crack, which is more robust than traditional back-propagation algorithm. It provides guidances and references for the maintenance of the gear drive system schemes.

Particle filtering based parameter estimation for systems with output-error type model structures

Abstract: The output-error model structure is often used in practice and its identification is important for analysis of output-error type systems. This paper considers the parameter identification of linear and nonlinear output-error models. A particle filter which approximates the posterior probability density function with a weighted set of discrete random sampling points is utilized to estimate the unmeasurable true process outputs. To improve the convergence rate of the proposed algorithm, the scalar innovations are grouped into an innovation vector, thus more past information can be utilized. The convergence analysis shows that the parameter estimates can converge to their true values. Finally, both linear and nonlinear results are verified by numerical simulation and engineering.

Enhanced Particle Filtering for Bearing Remaining Useful Life Prediction of Wind Turbine Drivetrain Gearboxes

Abstract: Bearing is the major contributor to wind turbine gearbox failures. Accurate remaining useful life prediction for drivetrain gearboxes of wind turbines is of great importance to achieve condition-based maintenance to improve the wind turbine reliability and reduce the cost of wind power. However, remaining useful life prediction is a challenging work due to the limited monitoring data and the lack of an accurate physical fault degradation model. The particle filtering method has been used for the remaining useful life prediction of wind turbine drivetrain gearboxes, but suffers from the particle impoverishment problem due to a low particle diversity, which may lead to unsatisfactory prediction results. To solve this problem, this paper proposes an enhanced particle filtering algorithm in which an adaptive neuro-fuzzy inference system is designed to learn the state transition function in the fault degradation model using the fault indicator extracted from the monitoring data; a particle modification method and an improved multinomial resampling method are proposed to improve the particle diversity in the resampling process to solve the particle impoverishment problem. The enhanced particle filtering algorithm is applied successfully to predict the remaining useful life of a bearing in the drivetrain gearbox of a 2.5 MW wind turbine equipped with a doubly-fed induction generator.

An improved particle filter for mobile robot localization based on particle swarm optimization

Abstract: As one of the most important issues in the field of mobile robotics, self-localization allows a mobile robot to identify and keep track of its own position and orientation as the robot moves through the environment. In this work, a hybrid localization approach based on the particle filter and particle swarm optimization algorithm is presented, focusing on the localization tasks when an a priori environment map is available. This results an accurate and robust particle filter based localization algorithm that is able to work in symmetrical environments. The performance of the proposed approach has been evaluated for indoor robot localization and compared with two benchmark algorithms. The experimental results show that the proposed method achieves robust and accurate positioning results in indoor environments, requiring fewer particles than the benchmark methods. This advance could be integrated in a wide range of mobile robot systems, helping to reduce the computational cost and improve the navigation efficiency.

Particle filtering-based recursive identification for controlled auto-regressive systems with quantised output

Abstract: Recursive prediction error method is one of the main tools for analysis of controlled auto-regressive systems with quantised output. In this study, a recursive identification algorithm is proposed based on the auxiliary model principle by modifying the standard stochastic gradient algorithm. To improve the convergence performance of the algorithm, a particle filtering technique, which approximates the posterior probability density function with a weighted set of discrete random sampling points is utilised to correct the linear output estimates. It can exclude those invalid particles according to their corresponding weights. The performance of the particle filtering technique-based algorithm is much better than that of the auxiliary model-based one. Finally, results are verified by examples from simulation and engineering.

Hybrid data-driven physics-based model fusion framework for tool wear prediction

Abstract: An integral part of modern manufacturing process management is to acquire useful information from machining processes to monitor machine and tool condition. Various models have been introduced to detect, classify, and predict tool wear, as a key parameter of the machining process. In more recent developments, sensor-based approaches have been attempted to infer the tool wear condition from real-time processing of the measurement data. Experiments show that the physics-based prediction models can include large uncertainties. Likewise, the measurement-based (or sensor-based) inference techniques are affected by sensor noise and measurement model uncertainties. To manage uncertainties and noise of both methods, a hybrid framework is proposed to fuse together the results of the prediction model and the measurement-based inference data in a stepwise manner. The fusion framework is an extension to the regularized particle filtering technique, used to facilitate updating the state prediction with a numerical inference model, when measurement models alone are not satisfactory. The results show significant improvement in tool wear state estimation, reducing the prediction errors by almost half, compared to the prediction model and sensor-based monitoring method used independently.

An Enhanced Mutated Particle Filter Technique for System State Estimation and Battery Life Prediction

Abstract: The particle filter (PF) has been used for the analysis of nonlinear, non-Gaussian dynamical systems with hidden state variables. However, PF has some limitations in real-world applications, for instant, the sample degeneracy and the impoverishment, which are considered as long-standing challenges in this research and development field. Although several techniques have been proposed in the literature for this purpose, they have some limitations: for example, they cannot represent the entire probability density function (pdf) effectively and are usually problem dependent. In this paper, an enhanced mutated PF (EMPF) technique is proposed to improve the performance of PFs. In the EMPF technique, first, a novel enhanced mutation approach is proposed to actively explore the posterior pdf to locate the high-likelihood area. Second, a new selection scheme is suggested to process low-weight particles for optimizing the posterior distribution and tackling sample degeneracy. Third, an outlier assessment method is adopted to monitor the overall pattern of the posterior distribution based on the interquartile range statistical analysis. The effectiveness of the proposed EMPF technique is verified by simulation tests. It is also implemented for the remaining useful life prediction of lithium-ion batteries. Test results show that the developed EMPF technique can capture a system’s dynamic behavior and track system characteristics effectively.

Tire-Stiffness and Vehicle-State Estimation Based on Noise-Adaptive Particle Filtering

Abstract: We present a novel approach to learning online the tire stiffness and vehicle state using only wheel-speed and inertial sensors. The deviations from nominal stiffness values are treated as a Gaussian disturbance acting on the vehicle. We formulate a Bayesian approach, in which we leverage particle filtering and the marginalization concept to estimate in a computationally efficient way the tire-stiffness parameters and the vehicle state. In the estimation model, the process and measurement noises are dependent on each other, and we present an efficient approach to account for the dependence. Our algorithm outperforms some previously reported approaches, both in terms of accuracy and robustness, and the results indicate significantly improved performance compared with a standard particle filter. Monte-Carlo trials on several experimental data sets verify that the estimator identifies the tire stiffness on both snow and dry asphalt within 1% on average, with a settling time of a few seconds. On snow, the largest steady-state error in any Monte-Carlo trial is less than 4%.

Remaining Useful Life Prediction of Lithium-Ion Batteries Using Neural Network and Bat-Based Particle Filter

Abstract: Predicting the remaining useful life (RUL) is an effective way to indicate the health of lithium-ion batteries, which can help to improve the reliability and safety of battery-powered systems. To predict the RUL, the line of research focuses on using the empirical degradation model followed by the particle filter (PF) algorithm, which is used for online updating the model’s parameters. However, this works well for specific batteries under specific discharge conditions. When the degradation trends cannot be presented by the chosen empirical model or the standard PF encounters impoverishment and degeneracy problem, the RUL prediction would be inaccurate. To improve the RUL prediction accuracy, we propose a novel approach by enhancing the existing method from two aspects. First, we introduce a neural network (NN) to model battery degradation trends under various operation conditions. As NN’s generalization and nonlinear representing ability, it outperforms the typical empirical degradation model. Second, the NN model’s parameters are recursively updated by the bat-based particle filter. The bat algorithm is used to move the particles to the high likelihood regions, which optimizes the particle distribution and thus reduces the degeneracy and impoverishment of PF. In this paper, quantitative evaluation is presented using two datasets with different batteries under different aging conditions. The results indicate that the proposed the approach can achieve higher RUL prediction accuracy than conventional empirical model and standard PF.

Behavioral Pedestrian Tracking Using a Camera and LiDAR Sensors on a Moving Vehicle.

Abstract: In this paper, we present a novel 2D–3D pedestrian tracker designed for applications in autonomous vehicles. The system operates on a tracking by detection principle and can track multiple pedestrians in complex urban traffic situations. By using a behavioral motion model and a non-parametric distribution as state model, we are able to accurately track unpredictable pedestrian motion in the presence of heavy occlusion. Tracking is performed independently, on the image and ground plane, in global, motion compensated coordinates. We employ Camera and LiDAR data fusion to solve the association problem where the optimal solution is found by matching 2D and 3D detections to tracks using a joint log-likelihood observation model. Each 2D–3D particle filter then updates their state from associated observations and a behavioral motion model. Each particle moves independently following the pedestrian motion parameters which we learned offline from an annotated training dataset. Temporal stability of the state variables is achieved by modeling each track as a Markov Decision Process with probabilistic state transition properties. A novel track management system then handles high level actions such as track creation, deletion and interaction. Using a probabilistic track score the track manager can cull false and ambiguous detections while updating tracks with detections from actual pedestrians. Our system is implemented on a GPU and exploits the massively parallelizable nature of particle filters. Due to the Markovian nature of our track representation, the system achieves real-time performance operating with a minimal memory footprint. Exhaustive and independent evaluation of our tracker was performed by the KITTI benchmark server, where it was tested against a wide variety of unknown pedestrian tracking situations. On this realistic benchmark, we outperform all published pedestrian trackers in a multitude of tracking metrics.

A parallel and robust object tracking approach synthesizing adaptive Bayesian learning and improved incremental subspace learning

Abstract: This paper presents a novel tracking algorithm which integrates two complementary trackers. Firstly, an improved Bayesian tracker(B-tracker) with adaptive learning rate is presented. The classification score of B-tracker reflects tracking reliability, and a low score usually results from large appearance change. Therefore, if the score is low, we decrease the learning rate to update the classifier fast so that B-tracker can adapt to the variation and vice versa. In this way, B-tracker is more suitable than its traditional version to solve appearance change problem. Secondly, we present an improved incremental subspace learning method tracker(S-tracker). We propose to calculate projected coordinates using maximum posterior probability, which results in a more accurate reconstruction error than traditional subspace learning tracker. Instead of updating at every time, we present a stop-strategy to deal with occlusion problem. Finally, we present an integrated framework(BAST), in which the pair of trackers run in parallel and return two candidate target states separately. For each candidate state, we define a tracking reliability metrics to measure whether the candidate state is reliable or not, and the reliable candidate state will be chosen as the target state at the end of each frame. Experimental results on challenging sequences show that the proposed approach is very robust and effective in comparison to the state-of-the-art trackers.

A Localized Adaptive Particle Filter within an Operational NWP Framework

Abstract: Particle filters are well known in statistics. They have a long tradition in the framework of ensemble data assimilation (EDA) as well as Markov chain Monte Carlo (MCMC) methods. A key chal...

Event-Trigger Heterogeneous Nonlinear Filter for Wide-Area Measurement Systems in Power Grid

Abstract: The development of distributed generation raises high requirement on accurate real-time state estimation with phasor measurement unit (PMU) for wide-area measurement systems (WAMSs). Although the particle filter is among the best in estimation performance, its computation burden is truly heavy toward smart sensors of current generation. Besides, the infrastructure for PMU data communication is confronted with the challenges of the amount of data boosted by the high sampling frequency of PMUs and booming grid size. This paper proposes an event-trigger heterogeneous nonlinear filter (ET-HNF) to solve these problems. A master–slave nonlinear filtering structure is designed for WAMS by taking full advantage of computation power both at generator node and estimation center. The local slave filter is established at generator node to perform the event-trigger strategy based on low computational unscented transformation. Moreover, a Monte Carlo-based filter served as the master, which improves estimation center’s filtering performance by cooperating with the center slave. Simulation results verify the feasibility and performance of ET-HNF using the standard IEEE 39-bus system with PMUs.

Positioning optimisation based on particle quality prediction in wireless sensor networks

Abstract: The particle degradation problem of particle filter (PF) algorithm caused by reduction of particle weights significantly influences the positioning accuracy of target nodes in wireless sensor networks. This study presents a predictor to obtain the particle swarm of high quality by calculating non-linear variations of ranging between particles and flags and modifying the reference distribution function. To this end, probability variations of distances between particles and star flags are calculated and the maximum inclusive distance using the maximum probability of high-quality particle swarm is obtained. The quality of particles is valued by the Euclidean distance between the predicted and real observations, and hereafter particles of high quality are contained in spherical coordinate system using the distance as diameter. The simulation results show that the proposed algorithm is robust and the computational complexity is low. The method can effectively improve the positioning accuracy and reduce the positioning error of target nodes.

MFBO-SSM: Multi-Fidelity Bayesian Optimization for Fast Inference in State-Space Models

Abstract: Nonlinear state-space models are ubiquitous in modeling real-world dynamical systems. Sequential Monte Carlo (SMC) techniques, also known as particle methods, are a well-known class of parameter estimation methods for this general class of state-space models. Existing SMC-based techniques rely on excessive sampling of the parameter space, which makes their computation intractable for large systems or tall data sets. Bayesian optimization techniques have been used for fast inference in state-space models with intractable likelihoods. These techniques aim to find the maximum of the likelihood function by sequential sampling of the parameter space through a single SMC approximator. Various SMC approximators with different fidelities and computational costs are often available for sample-based likelihood approximation. In this paper, we propose a multi-fidelity Bayesian optimization algorithm for the inference of general nonlinear state-space models (MFBO-SSM), which enables simultaneous sequential selection of parameters and approximators. The accuracy and speed of the algorithm are demonstrated by numerical experiments using synthetic gene expression data from a gene regulatory network model and real data from the VIX stock price index.

Hybrid SCA–TLBO: a novel optimization algorithm for global optimization and visual tracking

Abstract: A novel optimization algorithm called hybrid sine–cosine algorithm with teaching–learning-based optimization algorithm (SCA–TLBO) is proposed in this paper, for solving optimization problems and visual tracking. The proposed hybrid algorithm has better capability to escape from local optima with faster convergence than the standard SCA and TLBO. The effectiveness of this algorithm is evaluated using 23 benchmark functions. Statistical parameters are employed to observe the efficiency of the hybrid SCA–TLBO qualitatively, and results prove that the proposed algorithm is very competitive compared to the state-of-the-art metaheuristic algorithms. The hybrid SCA–TLBO algorithm is applied for visual tracking as a real thought-provoking case study. The hybrid SCA–TLBO-based tracking framework is used to experimentally measure object tracking error, absolute error, tracking detection rate, root mean square error and time cost as parameters. To reveal the capability of the proposed algorithm, a comparison of hybrid SCA–TLBO-based tracking framework and other trackers, viz. alpha–beta filter, linear Kalman filter and extended Kalman filter, particle filter, scale-invariant feature transform, particle swarm optimization and bat algorithm, is presented.

PoseRBPF: A Rao-Blackwellized Particle Filter for 6D Object Pose Tracking

Abstract: Tracking 6D poses of objects from videos provides rich information to a robot in performing different tasks such as manipulation and navigation. In this work, we formulate the 6D object pose tracking problem in the Rao-Blackwellized particle filtering framework, where the 3D rotation and the 3D translation of an object are decoupled. This factorization allows our approach, called PoseRBPF, to efficiently estimate the 3D translation of an object along with the full distribution over the 3D rotation. This is achieved by discretizing the rotation space in a fine-grained manner, and training an auto-encoder network to construct a codebook of feature embeddings for the discretized rotations. As a result, PoseRBPF can track objects with arbitrary symmetries while still maintaining adequate posterior distributions. Our approach achieves state-of-the-art results on two 6D pose estimation benchmarks. A video showing the experiments can be found at this https URL

Local-Diffusion-Based Distributed SMC-PHD Filtering Using Sensors With Limited Sensing Range

Abstract: We investigate the problem of distributed multitarget tracking by using a set of netted, collaborative sensors with limited sensing range (LSR), where each sensor runs a sequential Monte Carlo probability hypothesis density filter and exchanges relevant posterior information with its neighbors. The key challenge stems from the LSR of neighbor sensors whose fields of view (FoVs) are partially/non-overlapped, and therefore, they may observe different targets at the same time. With regard to the local common FoVs among neighbor sensors, the proposed distributed fusion approach, called local diffusion, performs one iteration of neighbor communication per filtering step in either of two means. One is given by immediate particle exchange, in which a reject-control operation is devised to reduce the number of communicating particles. The other is done by converting the particle distribution to Gaussian functions for parametric information exchange and fusion. The performance of both approaches has been experimentally investigated via simulation for different LSR situations and compared with cutting-edge approaches.

Improvement of maximum variance weight partitioning particle filter in urban computing and intelligence

Abstract: At present, urban computing and intelligence has become an important topic in the research field of artificial intelligence. On the other hand, computer vision as a crucial bridge between urban world and artificial intelligence is playing a key role in urban computing and intelligence. Conventional particle filter is derived from Karman filter, which theoretically based on Monte Carlo method. Sequential importance resampling (SIR) is implemented in conventional particle filter to avoid the degeneracy problem. In order to overcome the shortcomings of the resampling algorithm in the traditional particle filter, we proposed an optimized particle filter using the maximum variance weight segmentation resampling algorithm in this paper, which improved the performance of particle filter. Compared with the traditional particle filter algorithm, the experimental results show that the proposed scheme outperforms in terms of computational consumption and the accuracy of particle tracking. The final experimental results proved that the quality of the maximum variance weight segmentation method increased the accuracy and stability in motion trajectory tracking tasks.

Toward Accurate Vehicle State Estimation Under Non-Gaussian Noises

Abstract: Vehicle state including location and motion information plays an important role in various applications such as Internet of Vehicles (IoV), autonomous cars, and driving safety monitoring. Achieving accurate vehicle state is a challenging task in those applications due to the noise disturbances. Recent studies suggest that noise is not generally Gaussian distributed and many physical environments can be handled more accurately as non-Gaussian rather than Gaussian model. Inspired by this observation, we strive to improve the vehicle state estimation by investigating the effects of that assumption when process and measurement noises are non-Gaussian distributed. Here, process noise represents the noise during the state information processing. To that end, we exploit the generalized error distribution (GED) to compute the non-Gaussian probability density during the vehicle state estimation. We then derive extensive theoretical analysis targeting to estimate the parameters such as the mean and the variance (or covariance matrix) related to both process and measurement noises and reduce the computational burden of the distribution. Further, we propose a non-Gaussian particle filter for vehicle state estimation ( ${n}$ GPF-VSE) algorithm wherein we utilize the genetic operator resampling (GOR) technique to enhance the efficiency of particle filter (PF) relying on the selection of the importance sampling distribution. To evaluate the performance of the proposed approach, we conduct numerical simulations on the popular system of state-space equations and a real experiment for estimating the vehicle state. The results from the numerical simulations, experimental data and the statistical evaluation confirm that ${n}$ GPF-VSE outperforms existing methods in terms of vehicle state accuracy.

Magnetic-Based Indoor Localization Using Smartphone via a Fusion Algorithm

Abstract: The detection of indoor location based on magnetic fields collected by embedded sensors in smartphones has been progressed rapidly. Most current approaches rely on the particle filter (PF) scheme which combines the pedestrian dead reckoning (PDR) technique with patterns of previously recorded magnetic field intensity. The key challenges include inherent blindness and particle degradation problems. Here, a fusion algorithm combining the extended Kalman filter (EKF) and the PF scheme is proposed to address these issues. EKF is first used to reduce the possible location regions by fusing the PDR and magnetic field intensity results. The particle generation, update, and resampling processes are conducted afterward, and the final position is calculated by using the weighted mean of particles. As such, the blindness and particle degradation problems are alleviated by using particles in the reduced location regions at each processing step. Experiments show a localization accuracy of 1-2m when the user walks smoothly, which is better than those of traditional PF schemes, especially in cases under heavy magnetic distortions by using a reduced number of particles.

A Particle Filter-Based Reinforcement Learning Approach for Reliable Wireless Indoor Positioning

Abstract: Positioning is envisioned as an essential enabler of future fifth generation (5G) mobile networks due to the massive number of use cases that would benefit from knowing users’ positions. In this work, we propose a particle filter-based reinforcement learning (PFRL) approach for the robust wireless indoor positioning system. Our algorithm integrates information of indoor zone prediction, inertial measurement units, wireless radio-based ranging, and floor plan into an particle filter. The zone prediction method is designed with an ensemble learning algorithm by integrating individual discriminative learning methods and Hidden Markov Models. Further, we integrate the particle filter approach with a reinforcement learning-based resampling method to provide robustness against localization failure problems such as the kidnapping robot problem. The PFRL approach is validated on a two-tier architecture, in which distributed machine learning tasks are hosted at client and edge layer. Experiment results show that our system outperforms traditional terminal-based approaches in both stability and accuracy.