Particle Filter Framework Research Articles

Tightly coupled integration of vector HD map, LiDAR, GNSS, and INS for precise vehicle navigation in GNSS-challenging environment

ABSTRACT High-precision positioning and navigation are necessary for autonomous driving. GNSS RTK and INS integrated system is commonly used in vehicular navigation. But it suffers from severe signal reflections and blockages of GNSS signals and error accumulation of INS with MEMS-IMU in GNSS-challenging environment. In addition, high-definition (HD) maps in vector format and light detection and ranging (LiDAR) are two common options for intelligent vehicles. A tightly coupled localization method with vector HD map, LiDAR, GNSS RTK, and INS is proposed to take advantage of their complementary characteristics to accurately navigate a vehicle in a GNSS-challenging environment. The method is based on a particle filter (PF) framework. Lateral positions of particles are constrained by LiDAR measurements and lane information in the vector HD map. A constrained and damped LAMBDA searching method is proposed to update particle weights, aiming to find accurate longitudinal localization. Experimental results prove that our method can maintain sub-meter level horizontal positioning accuracy in GNSS-challenging environment, with improvements of (77%, 75%, 64% and 65%), regarding to three-dimensional position and yaw, compared to traditional GNSS-RTK/INS integration, while the improvement of the popular GNSS-RTK/INS/LiDAR integrated framework LIO-SAM is (16%, 53%, 48% and 51%).

Hunt-inspired Transformer for visual object tracking

This paper presents a hunt-inspired Transformer for visual object tracking, dubbed as HuntFormer. The HuntFormer focuses on robust target detection and identification, simulating natural hunting processes. Specifically, the HuntFormer comprises two essential module designs including a predictor for detection and a verifier for identification. The predictor emulates the detection stage by designing a motion trajectory guided particle filter, which identifies potential target locations by predicting the motion state within a particle filtering framework. The predictor utilizes spatio-temporal correlation scores between dynamic target templates and the search region to guide the learning process to generate a set of reliable particles. This enables the base tracker to narrow its search range to focus on the target, and swiftly re-detect the target in case of model drift. Once the target is re-detected, the verifier assesses the detection result as a reliable tracked item. The verifier initially maintains a dynamic memory that stores reliable target templates and their corresponding locations in the motion trajectory. It then models the uncertainty of appearance information within this memory probabilistically. The output uncertainty score determines whether the memory gets updated or not. Ultimately, the predictor and the verifier collaborate, ensuring a robust tracking outcome. Extensive evaluations on six challenging benchmark datasets demonstrate HuntFormer’s favorable performance against various state-of-the-art trackers. Notably, in the VOT-LT2022 tracking challenge, the HuntFormer won the third place with an F-score of 0.598, closely competing for the second place with an F-score of 0.600.

Reliable urban vehicle localization under faulty satellite navigation signals

Reliable urban navigation using global navigation satellite system requires accurately estimating vehicle position despite measurement faults and monitoring the trustworthiness (or integrity) of the estimated location. However, reflected signals in urban areas introduce biases (or faults) in multiple measurements, while blocked signals reduce the number of available measurements, hindering robust localization and integrity monitoring. This paper presents a novel particle filter framework to address these challenges. First, a Bayesian fault-robust optimization task, formulated through a Gaussian mixture model (GMM) measurement likelihood, is integrated into the particle filter to mitigate faults in multiple measurement for enhanced positioning accuracy. Building on this, a novel test statistic leveraging the particle filter distribution and the GMM likelihood is devised to monitor the integrity of the localization by detecting errors exceeding a safe threshold. The performance of the proposed framework is demonstrated on real-world and simulated urban driving data. Our localization algorithm consistently achieves smaller positioning errors compared to existing filters under multiple faults. Furthermore, the proposed integrity monitor exhibits fewer missed and false alarms in detecting unsafe large localization errors.

An electrochemical-thermal coupling model for lithium-ion battery state-of-charge estimation with improve dual particle filter framework

Lithium-ion battery state-of-charge (SOC) estimation plays an indispensable role in the battery management system functionalities. To achieve SOC estimation with high accuracy, a precise battery model is required. Among various battery models, electrochemical model (EM) stands out for its ability to provide insights into the electrochemical mechanisms of the battery. However, the model parameters will vary with the battery running and aging, besides, the model-based methods cannot well handle the model errors. Therefore, to ensure estimation accuracy, selecting a more insightful battery model and a more effective estimation algorithm becomes crucial. This paper proposes an electrochemical model-based SOC estimation method with online parameter adjustment and an optimized algorithm. Firstly, a single particle model (SPM) with electrolyte and thermal dynamics is introduced and simplified by an approximate solution. Secondly, the particle swarm optimization (PSO) method is used to solve the problem of the particle filter (PF) algorithm, such as particle degradation and particle diversity decrease. Finally, during SOC estimating, a double-scale dual particle filter (D-PF) is utilized to adjust parameters as battery aging and temperature changes. The battery tests are applied under different working conditions and temperatures. Experimental results show that the error of SOC estimation reduces significantly compared to the previous electrochemical model-based method, and the proposed method has a higher accuracy and effectiveness.

Hybrid prognostics to estimate cutting inserts remaining useful life based on direct wear observation

The aim of this study is to monitor tool wear through recurrent direct observation, to automatically and optimally assess when it is really necessary to change tools. To achieve this goal, a hybrid prognosis algorithm is formulated to estimate cutting tools’ remaining useful life. Since cutting speed, and more generally process parameters, influences the rate of tool degradation, an adaptive prognosis strategy is presented on the basis of flank wear measurements through the application of a Particle Filter (PF) framework. The adaptability feature allows tracking changes in flank wear evolution. The idea is to fit available degradation curves of cutting tool flank land measurements, through the use of data-driven models, i.e. Multi-Layer Perceptrons (MLP) and cubic polynomials (P3). The Remaining Useful Life of the cutting tool is estimated together with its probability density function, by using a PF framework to adapt MLP weights (or P3 coefficients) along with online flank wear measurements. The devised algorithm was proven to adapt to wear trends from the field, obtained with cutting parameters not previously tested, making it suitable for a robust implementation. The approach was tested when trained upon a single run-to-failure, and validated upon four run-to-failures in different cutting conditions according to a cross-validation inspired technique. P3 was found to be more reliable (from the metrics perspective), whereas MLP allowed to be accurate with greater advance, offering a practical advantage. The proposed algorithm may also be adapted to integrate physical features, like specific force coefficients, with direct wear measurements.

An integrated fuse information fusion method for infrared imaging detector and range finder based on improved particle filter

AbstractAiming at the traditional integrated fuse information fusion method that cannot simultaneously obtain high‐precision burst control parameters such as burst delay time, azimuth angle, and detection angle, the infrared imaging detector and range finder information fusion method based on particle filtering with BAS algorithm is proposed. Under the centralized information fusion structure that can obtain higher accuracy, first, use a faster linear interpolation method to realize the time alignment of the two asynchronous sensors. Second, for the strong non‐linear observation model in the missile target rendezvous model, use a particle filter framework that is not affected by the nonlinearity of the system function and non‐gaussian noise. Finally, Utilizing the single search BAS algorithm with low computational complexity to solve the problem of particle degradation in particle filtering and improve the accuracy of burst control. The experiment shows that in a typical simulation environment, the optimized information fusion method improves the burst delay time accuracy by 61.82%, azimuth accuracy by 66.28%, detection angle accuracy by 36.15%, and simulation time is shorter compared to traditional methods.

MODEL ERROR ESTIMATION USING PEARSON SYSTEM WITH APPLICATION TO NONLINEAR WAVES IN COMPRESSIBLE FLOWS

In data assimilation, the description of the model error uncertainty is of the utmost importance because, when incorrectly defined, it may lead to information loss about the real state of the system. In this work, we proposed a novel approach that finds the optimal distribution for describing the model error uncertainty within a particle filtering framework. The method was applied to nonlinear waves in compressible flows. We investigated the influence of observation noise statistics, resolution of the numerical model, smoothness of the solutions, and sensor location. The results showed that in almost all situations the Pearson Type I is preferred, but with different curve-shape characteristics, namely, skewed, nearly symmetric, ∩-, ∪-, and J-shaped. The distributions became, in most cases, ∪-shaped when the sensors were located near the discontinuities.

A robust particle filter for ambiguous updates of underwater terrain-aided navigation

This paper proposes a novel method to improve the robustness of particle filtering (PF) in relatively flat terrain that lack sufficient features for accurate positioning. In terrain -aided navigation (TAN), similar terrain profiles lead to multimodality in the posterior distribution. It not only reduces the accuracy of current position estimation, but also affects the subsequent estimation of PF, and even causes filter divergence. When no additional information is available, the historical estimation is employed to correct ambiguous updates caused by multimodal posteriors. First, a strategy for identifying ambiguous updates is proposed by analyzing the particle set distribution using the clustering method and covariance. Inspired by out-of-sequence measurement (OOSM), once the ambiguous updates are detected, the ambiguous estimates are corrected by introducing high-quality previous information. Moreover, an efficient solution is provided for the storage and computation requirements of OOSM within the PF framework. To verify the effectiveness of the proposed algorithm, simulation and experimental validation are designed. By comparing with PF, mixture particle filtering (MPF), and OOSMPF algorithms, the proposed algorithm demonstrates better estimation accuracy and robustness in terrain flat areas.

A Degradation Empirical-Model-Free Battery End-of-Life Prediction Framework Based on Gaussian Process Regression and Kalman Filter

Predicting the battery’s end-of-life (EOL) with uncertainty quantification is critical for ensuring system safety and reliability. This paper presents a hybrid framework for battery EOL prediction and its uncertainty assessment based on Gaussian process regression (GPR) and Kalman filter (KF). First, a KF-based empirical-model-free state tracking phase is applied for the available partial battery degradation data. Then, the original time series forecasting problem of degradation curves is converted to the prediction of the virtual degradation rate and acceleration. Next, the prediction of the virtual degradation rate and acceleration is executed by the iterative GPR multi-step ahead prediction strategy with moving sliding windows. Finally, the uncertainty assessment is carried out based on the sliding window length determination process. The effectiveness of our proposed method is validated on the open-source lithium-ion battery degradation dataset from the University of Oxford. Extensive EOL prediction tests have been carried out from 40% (early-stage), 60% (middle-stage), and 80% (late-stage) of the dataset, respectively. Compared with the popular EOL prediction method within particle filter framework, the predicted mean EOL cycle by our method is closer to the true value with a smaller range of prediction uncertainty.

Deep Neural Network-Based Fusion Localization Using Smartphones

Indoor location-based services (LBS) have tremendous practical and social value in intelligent life due to the pervasiveness of smartphones. The magnetic field-based localization method has been an interesting research hotspot because of its temporal stability, ubiquitousness, infrastructure-free nature, and good compatibility with smartphones. However, utilizing discrete magnetic signals may result in ambiguous localization features caused by random noise and similar magnetic signals in complex symmetric and large-scale indoor environments. To address this issue, we propose a deep neural network-based fusion indoor localization system that integrates magnetic and pedestrian dead reckoning (PDR). In this system, we first propose a ResNet-GRU-LSTM neural network model to achieve magnetic localization more accurately. Afterward, we put forward a multifeatured-driven step length estimation. A hierarchy GRU (H-GRU) neural network model is proposed, and a multidimensional dataset using acceleration and a gyroscope is constructed to extract more valid characteristics. Finally, more reliable and accurate pedestrian localization can be achieved under the particle filter framework. Experiments were conducted at two trial sites with two pedestrians and four smartphones. Results demonstrate that the proposed system achieves better accuracy and robustness than other traditional localization algorithms. Moreover, the proposed system exhibits good generality and practicality in real-time localization with low cost and low computational complexity.

Hypothesis Scoring for Confidence-Aware Blood Pressure Estimation With Particle Filters.

We propose our Confidence-Aware Particle Filter (CAPF) framework that analyzes a series of estimated changes in blood pressure (BP) to provide several true state hypotheses for a given instance. Particularly, our novel confidence-awareness mechanism assigns likelihood scores to each hypothesis in an effort to discard potentially erroneous measurements - based on the agreement amongst a series of estimated changes and the physiological plausibility when considering DBP/SBP pairs. The particle filter formulation (or sequential Monte Carlo method) can jointly consider the hypotheses and their probabilities over time to provide a stable trend of estimated BP measurements. In this study, we evaluate BP trend estimation from an emerging bio-impedance (Bio-Z) prototype wearable modality although it is applicable to all types of physiological modalities. Each subject in the evaluation cohort underwent a hand-gripper exercise, a cold pressor test, and a recovery state to increase the variation to the captured BP ranges. Experiments show that CAPF yields superior continuous pulse pressure (PP), diastolic blood pressure (DBP), and systolic blood pressure (SBP) estimation performance compared to ten baseline approaches. Furthermore, CAPF performs on track to comply with AAMI and BHS standards for achieving a performance classification of Grade A, with mean error accuracies of -0.16 ± 3.75 mmHg for PP (r = 0.81), 0.42 ± 4.39 mmHg for DBP (r = 0.92), and -0.09 ± 6.51 mmHg for SBP (r = 0.92) from more than test 3500 data points.

Ensemble Remaining Useful Life Prediction for Lithium-Ion Batteries With the Fusion of Historical and Real-Time Degradation Data

Remaining useful life (RUL) prediction is a critical task in prognostics and health management. The performances of traditional RUL prediction approaches for lithium-ion batteries are usually affected by the uncertainties involved in the data analysis and model selection. This paper proposes an ensemble prognostic approach under the particle filter (PF) framework to improve the prediction accuracy in consideration of the uncertainties. In PF algorithm, an optimal weights initialization method is proposed with the comprehensive consideration of model bias and variance, and a novel weighting scheme is proposed to optimize the ensemble model performance by assigning time-varying and degradation-dependent weights with the fusion of historical and real-time degradation data. Besides, a data noise quantification method is proposed and applied in the PF algorithm to solve the hyperparameter setting problem. The effectiveness of the proposed approach is illustrated through the real datasets obtained from two types of lithium-ion batteries.

Physics-based model informed smooth particle filter for remaining useful life prediction of lithium-ion battery

Accurately predicting the remaining useful life (RUL) of lithium-ion batteries is essential for battery management systems (BMS) as rapid capacity declines and failure can impact equipment operation and pose safety hazards. However, battery aging is a complex electrochemical process influenced by various factors such as cycle time, temperature, and loading conditions. To address this, a physics-informed smooth particle filter (SPF) framework for RUL prediction is proposed in this work, which estimates parameters of a single particle model (SPM) of LiBs by extracting three main degradation mechanisms: active material loss in positive and negative electrodes and loss of lithium inventory. Unlike traditional prognostic frameworks, this approach utilizes the SPM to estimate degradation parameters directly from voltage and capacity data, enabling more accurate quantification of degradation mechanisms and prediction of capacity fade trends. The estimated capacity is then used to develop an RUL predictor based on an SPF, which produces more accurate RUL predictions compared to conventional capacity-based methods. The proposed framework achieves a best-case RUL prediction of 2402 cycles at the prediction starting point of 2000 cycles, with a minimum relative error of around 0.089% compared to approximately 0.8% for the traditional framework. Additionally, the proposed framework is demonstrated to be dependable and robust, even when dealing with LiBs data containing Gaussian white noise and dynamic discharging profiles.

Real-Time Indoor Positioning Based on BLE Beacons and Pedestrian Dead Reckoning for Smartphones

Nowadays, smartphones have become indispensable in people’s daily work and life. Since various sensors and communication chips have been integrated into smartphones, it has become feasible to provide indoor positioning using phones. This paper proposes such a solution based on a smartphone, combining Bluetooth low energy (BLE) and pedestrian dead reckoning (PDR) in the particle filter framework to realize real-time and reliable indoor positioning. First, the smartphone’s built-in accelerometer, magnetometer, and gyroscope are used to provide data measurements and formulate a feasible method for PDR. Second, a range-free weighted centroid algorithm is proposed to realize BLE-based localization with low computation complexity. However, a single positioning technology has limitations, e.g., the cumulative error of PDR and the received signal strength fluctuation of BLE. Finally, to exploit the complementary strengths of each technology, a fusion framework utilizing a particle filter is proposed to combine PDR and BLE-based methods and provides more stable and accurate positioning results. Experiments are conducted on a floor in a campus building. Experimental results show that our proposed fused positioning method offers more accurate and stable performance in the long run compared with single PDR or BLE-based positioning. The achieved average positioning error is 1.34 m, which is reduced by 24.16% compared with PDR positioning and 10.60% compared with BLE-based positioning. Moreover, about 95% of the positioning errors are smaller than 1.7 m. The proposed fused positioning method has a vast application prospect in indoor navigation, indoor user tracking, and interactive experience for indoor visitors.

Prediction of multiple fatigue crack growth based on modified Paris model with particle filtering framework

This paper presents a novel method based on modified Paris model with particle filtering (PF) framework to prognosis fatigue multi-cracks in metal structures. Since conventional Paris model is only applicable for the single crack, modified Paris model is proposed by introducing the mutual interaction between multiple cracks into stress intensity factor (SIF) expression and utilized to portray multiple fatigue crack growth process. To accurately monitor multi-cracks lengths, Lamb waves are periodically acquired during the fatigue growth of multiple cracks and fed into a deep autoencoder (DAE) network to automatically track response signals variations. Online monitoring model of multi-cracks lengths is then constructed by fitting the mapping relationship between the deep damage feature extracted by the bottleneck layer of DAE network and crack length. PF framework is further adopted to strategically fusing the prediction results of modified Paris model and real-time measurements of online monitoring model to reduce the uncertainties of material parameters and obtain more reliable prognosis results of multi-cracks lengths. The proposed method is demonstrated on center-hole metal specimens with two fatigue cracks. Experimental results show that the proposed method can accurately prognosis two fatigue cracks lengths, in which modified Paris model could more accurately describe multiple fatigue crack growth than conventional Paris model, and online monitoring model based on deep damage feature could more accurately track the fatigue crack length than artificial damage feature.

Observation Model for Robust Localization of Mobile Robots based on Particle Filter in Changing Environments

For autonomous mobile robot navigation, localization is an essential capability. Given a mobile robot equipped with a 3D LiDAR sensor, an environment map composed of point cloud is built beforehand. The robot is thus allowed to localize the position in the map using the sensor scan data. However, the environment sometimes changes due to obstacles. Under the changing environment, the localization capability of the robot might be decreased. For this challenge, we propose a sensor observation model in a framework of particle filter based localization. In the observation model, we focus on the distance and distribution of point clouds of the map and sensor scan data. In the experiments, a mobile robot is moved by an operator in a virtual environment with obstacles. The robot based on the proposed observation model is able to localize the position in both the original and changing environments with the same accuracy. From the results, we finally show the robustness of the localization capability for changing environments.

Long Baseline Acoustic Localization Based on Track-Before-Detect in Complex Underwater Environments

This paper considers the long baseline (LBL) acoustic localization problem. For this problem, traditional two-step localization methods first estimate the time of arrivals (TOAs) of the target signal and then solve the target’s position based on the estimated TOAs. Since TOA estimation is performed at each buoy independently, they are suboptimal without considering the physical basis that the TOAs estimated from different buoys all correspond to the same target position. To overcome this drawback, we propose a novel localization approach based on track-before-detect (TBD) and particle filtering (PF) theory, which directly determines the target’s position. Since this method breaks from the traditional two-step paradigm, several extra challenges arising from the two-step paradigm itself can be avoided. Specifically, we make two main contributions. First, a unique method is given for designing the likelihood function of the PF framework, which defines the likelihood function as the product of the matched filter outputs of multiple buoys conditioned on the particle states. As this design scheme is without any assumptions on the emission signal type, it can be applied to any signal type. Second, multiple models are employed to sample the particles to reduce the performance loss caused by target maneuvers. Additionally, an auxiliary variable is used to improve the sampling efficiency of the multiple models. The efficacy of the proposed method is demonstrated both in simulations and on real datasets. Experimental results show that the proposed algorithm outperforms the traditional LBL localization methods under harsh conditions.

Region Probability Map-Guided Fast Wide-Area Multiobject Detection

Detecting multiobjects in a wide-area scenario efficiently is a critical technology for industry and security applications. The detection performance has benefited enormously from the probability estimation of the unknown environment. Representative methods, such as particle filter (PF), construct the probability distribution model and iteratively locate target objects. However, randomly sampling and detecting image patches in the large covering field-of-view (FOV) make these methods inefficient and computation costly. To address these issues, we propose a region probability map (RPM)-guided fast wide-area detection system that can simultaneously detect multiobjects from a large FOV at 300 frames per second (fps) through a coarse-to-fine-grained detection paradigm. Specifically, a segmentation-based RPM generation module is introduced to assign probability measurements to different regions of the panoramic image, which models how likely the desired objects will occur in these regions. Then, based on the generated probability map of the whole scene, a novel RPM-guided PF framework is proposed to speed up the detection process by concentrating the detection on high-probability areas. Finally, a rapid and low-latency active detection system based on a wide-angle camera, a high-speed camera, and an ultrafast galvano mirror is implemented, which gains a 15.38% efficiency improvement while achieving more accurate detection compared with existing methods. Extensive experimental results verify the robustness and effectiveness of our proposed system.

Improving the particle filter for data assimilation in hydraulic modeling by using a Cauchy likelihood function

Over the recent decades, particle filter (PF)-based data assimilation has been adopted to improve hydrodynamic simulation due to its wide applicability and effectiveness for nonlinear and non-Gaussian models. The particle weighting is the core procedure to combine the observations and simulations within the PF framework, in which the Gaussian likelihood function (GLF) is commonly used to update weight values. However, there are some widely acknowledged challenges within the GLF-based PF. Among them, the excessive dispersion problem occurs when the standard deviation of the Gaussian likelihood is small, which may even cause filtering failures. Unfortunately, this issue has not been sufficiently addressed in the hydrology community. More attention needs to be paid to the likelihood function since the errors of the observed water stage (represented by the standard deviation) from hydrometric stations are usually small. This study focuses on the form of the likelihood function and parameter tuning within the particle weighting procedure of the PF for data assimilation. The limitations of using the common GLF were clarified and a novel Cauchy likelihood function (CLF) was proposed and tested in both a synthetic and a real-world case study using a PF-based hydraulic model. Comparisons showed that the CLF-based PF can effectively relieve the problem of excessive dispersion within the GLF-based PF framework, resulting in more stable and more accurate results. These findings not only provide another option for the selection of likelihood function within the PF framework, but also offer reliable evidence for improving hydraulic modeling by incorporating high-precision observations using the data assimilation technique.

Predicting Model for Air Transport Demand under Uncertainties Based on Particle Filter

The outbreak of the COVID-19 has brought about huge economic loss and civil aviation industries all over the world have suffered severe damage. An effective method is urgently needed to accurately predict air-transport demand under the influences of such accidental factors. This paper proposes a novel predicting framework for the air-transport demand considering the uncertainties caused by accidental factors including regional wars, climatic anomalies, and virus outbreaks. By employing a seasonal autoregressive integrated moving average (sARIMA) model as the basic model, a particle filter (PF)-based sARIMA-pf model is proposed. The applicability of adapting the high-order sARIMA model as the state transition model in a PF framework is shown and proven to be effective. The proposed method has the advantage of coping with short-term prediction with known uncertainties. By conducting case studies on the prediction of air passenger traffic volume in China, the sARIMA-pf model showed better performance than the sARIMA model and improved the accuracy by 49.29% and 44.96% under the conventional and pandemic scenarios, respectively, when using the root mean square error (RMSE) as the indicator.