Data Fusion Of Measurements Research Articles

Real-Time Autonomous Vehicle Localization Based on Particle and Unscented Kalman Filters

In this paper, a real-time Monte Carlo localization (RT_MCL) method for autonomous cars is proposed. Unlike the other localization approaches, the balanced treatment of both pose estimation accuracy and its real-time performance is the main contribution. The RT_MCL method is based on the fusion of lidar and radar measurement data for object detection, a pole-like landmarks probabilistic map and a tailored particle filter for pose estimation. The lidar and radar are fused using the unscented Kalman filter (UKF) to provide pole-like static-object pose estimations that are well suited to serve as landmarks for vehicle localization in urban environments. These pose estimations are then clustered using the grid-based density-based spatial clustering of applications with noise algorithm to represent each pole landmark in the form of a source-point model to reduce computational cost and memory requirements. A reference map that includes pole landmarks is generated offline and extracted from a 3-D lidar to be used by a carefully designed particle filter for accurate ego-car localization. The particle filter is initialized by the fused GPS + IMU measurements and used an ego-car motion model to predict the states of the particles. The data association between the estimated landmarks by the UKF and that in the reference map is performed using the iterative closest point algorithm. The RT_MCL is implemented using the high-performance language C++ and utilizes highly optimized math and optimization libraries for best real-time performance. Extensive simulation studies have been carried out to evaluate the performance of the RT_MCL in both longitudinal localization and lateral localization.

Effective Passive Multitarget Localization Using Maximum Likelihood

Localization of multiple targets is a challenging task due to immense complexity regarding data fusion received at the sensors. In this context, we propose an algorithm to solve the problem for an unknown number of emitters without prior knowledge to address the data fusion problem. The proposed technique combines the time difference of arrival (TDOA) and frequency difference of arrival (FDOA) measurement data fusion which further uses the maximum likelihood of the measurements received at each sensor of the surveillance region. The measurement grids of the sensors are used to perform data association. The simulation results show that the proposed algorithm outperforms the multipass grid search and further effectively eliminated the ghost targets created due to the fusion of measurements received at each sensor. Moreover, the proposed algorithm reduces the computational complexity compared to other existing algorithms as it does not use repeated steps for convergence or any biological evolutions. Furthermore, the experimental testing of the proposed technique was executed successfully for tracking multiple targets in different scenarios passively.

Real-time lidar and radar fusion for road-objects detection and tracking

In this paper, a real-time road-object detection and tracking (LR_ODT) method for autonomous driving is proposed. The method is based on the fusion of lidar and radar measurement data, where they a...

Kalman-filter-based sensor fusion applied to road-objects detection and tracking for autonomous vehicles

In this article, a real-time road-Object Detection and Tracking (LR_ODT) method for autonomous driving is proposed. This method is based on the fusion of lidar and radar measurement data, where they are installed on the ego car, and a customized Unscented Kalman Filter is employed for their data fusion. The merits of both devices are combined using the proposed fusion approach to precisely provide both pose and velocity information for objects moving in roads around the ego car. Unlike other detection and tracking approaches, the balanced treatment of both pose estimation accuracy and its real-time performance is the main contribution in this work. The proposed technique is implemented using the high-performance language C++ and utilizes highly optimized math and optimization libraries for best real-time performance. Simulation studies have been carried out to evaluate the performance of the LR_ODT for tracking bicycles, cars, and pedestrians. Moreover, the performance of the Unscented Kalman Filter fusion is compared to that of the Extended Kalman Filter fusion showing its superiority. The Unscented Kalman Filter has outperformed the Extended Kalman Filter on all test cases and all the state variable levels (−24% average Root Mean Squared Error). The employed fusion technique shows how outstanding is the improvement in tracking performance compared to the use of a single device (−29% Root Mean Squared Error with lidar and −38% Root Mean Squared Error with radar).

Data fusion of citizen-generated smartphone discharge measurements in Tunisia

Water resources management techniques have been evolving over the years, introducing new ways of monitoring and collecting data that improve both the quality and quantity of water-related information. Among them, Citizen Science (CS) has been introduced in the field of environmental monitoring as a novel approach that involves a direct collaboration between citizens, scientists and local authorities.In Tunisia, the lack of reliable hydrological data about rivers’ discharge remains a major issue, despite the recent governmental efforts to reinforce the existing official monitoring systems. In this study, we show how this problem can be efficiently addressed using a CS methodology. Citizens at two locations in Tunisia (Slouguia and Medjez-N5) have monitored discharge of the Medjerda river (the most important water resource in the country) using a mobile phone application for a series of hydrological events. Using a Best Linear Unbiased Predictor (BLUP) as a data fusion procedure for combining the various CS measurements, the results show that predicted discharge at both locations is in very good agreement with the reference data collected for the same hydrological events. Based on a variance decomposition, this approach allows us as well to properly assess the respective part of the random errors that are related to the citizens and measuring devices.It is concluded that CS-based discharge data collection is a promising cost-effective way for obtaining reliable and numerous measurements. The monitoring approach based on the use of a mobile phone application is thus quite valuable for complementing the existing Tunisian monitoring system, as well as for empowering local communities. Furthermore, this approach can be applied at larger scales in the country by involving more citizens and adding other sites, which should support the national efforts for better and smarter water resources management.

Road-objects tracking for autonomous driving using lidar and radar fusion

Abstract In this paper, based on the fusion of Lidar and Radar measurement data, a real-time road-Object Detection and Tracking (LR_ODT) method for autonomous driving is proposed. The lidar and radar devices are installed on the ego car, and a customized Unscented Kalman Filter (UKF) is used for their data fusion. Lidars are accurate in determining objects’ positions but significantly less accurate on measuring their velocities. However, Radars are more accurate on measuring objects velocities but less accurate on determining their positions as they have a lower spatial resolution. Therefore, the merits of both sensors are combined using the proposed fusion approach to provide both pose and velocity data for objects moving in roads precisely. The Grid-Based Density-Based Spatial Clustering of Applications with Noise (GB-DBSCAN) clustering algorithm is used to detect objects and estimate their centroids from the lidar and radar raw data. Then, the estimation of the object’s velocity as well as determining its corresponding geometrical shape is performed by the RANdom SAmple Consensus (RANSAC) algorithm. The proposed technique is implemented using the high-performance language C++ and utilizes highly optimized math and optimization libraries for best real-time performance. The performance of the UKF fusion is compared to that of the Extended Kalman Filter fusion (EKF) showing its superiority. Simulation studies have been carried out to evaluate the performance of the LR ODT for tracking bicycles, cars, and pedestrians.

A Novel False Data Injection Attack Detection Model of the Cyber-Physical Power System

The False data injection attack (FDIA) against the Cyber-Physical Power System (CPPS) is a kind of data integrity attack. With more and more cyber vulnerabilities detected out, different types of FDIAs are emerging as severe threats to the stable operation of CPPS gradually. In this paper, the invasion pathway of the FDIA against CPPS is explored in detail, and a novel FDIA detection model based on ensemble learning is further provided. First, a pseudo-sample database is built to assist the training and evaluation of this model, and it's more important to update the model in the future. Furthermore, the optimal feature set is extracted to characterize the behavior of the FDIA, which improves the precision of the FDIA detection model. Finally, a focal-loss-lightgbm (FLGB) ensemble classifier is constructed to detect the FDIA behavior automatically and accurately. We illustrated the performance of this model by a fusion of measurement data and power system audit logs. This model utilizes the offline training way, the conclusion shows the high precision and stability of this model, which ensures the stable operation of the smart grid and improves the FDIA resistance ability of the CPPS.

Comparison of sixteen methods for fusion of data from impulse-radar sensors and depth sensors applied for monitoring of elderly persons

This paper is devoted to the comparison of sixteen methods for fusion of measurement data from impulse-radar sensors and infrared depth sensors, i.e. two sensor technologies that may be employed in care services for elderly persons. These methods are compared with respect to their potential for decreasing the uncertainty of estimation of monitored person’s position: eight of them consist in fusing the impulse-radar data and depth data whenever new data points are available, and the other eight consist in fusing the whole sequences of the data acquired during a predefined time interval. The numerical experiments, based on the real-world data, show that the best overall results are obtained for two methods of data fusion, viz. a method based on the Kalman filter and a method using the Tikhonov regularization technique to generate a smooth approximation of the data.

Application of artificial neural networks for fusion of data from radar and depth sensors applied for persons’ monitoring

The research reported in this paper is related to the fusion of measurement data from the impulse-radar sensors and infrared depth sensors applied in a system for unobtrusive monitoring of elderly persons. Three methods for data fusion, based on the artificial neural networks – one trained on real-world data, and two trained on synthetic data generated on the basis of two different models of the data – are compared with respect to their capacity of decreasing the uncertainty of position estimation in a series of experiments which involved the tracking of a moving person.

Monitoring crop water content for corn and soybean fields through data fusion of MODIS and Landsat measurements in Iowa

Vegetation water content (VWC) is of vital significance to many applications in agriculture, such as crop yield estimation and precision irrigation. However, previous studies mainly estimate VWC at either high spatial resolution or temporal resolution due to the limitation of space-borne observation systems. In this paper, we target on monitoring daily plant VWC as well as canopy VWC at 30 m high spatial resolution with the fusion of optical measurements from the Land Remote Sensing Satellite (Landsat) Operational Land Imager (OLI) and the Moderate Resolution Imaging Spectroradiometer (MODIS) in an agricultural area. The study area locates in the central of Iowa, the United States (USA), the study period ranges from August 1 to August 17, 2016. Landsat 8 OLI observations are with 16-day revisit cycle and 30-m spatial resolution, MODIS daily surface reflectance product used in this study has 500-m spatial resolution. Six Landsat OLI images and thirty MODIS images were used to estimate daily remotely sensed reflectance data at 30m first. Then, the fused satellite measurements were used to calculate Normalized Difference Water Index (NDWI). Ground-based plant VWC and canopy VWC measurements were collected during the Soil Moisture Active Passive Validation Experiment (SMAPVEX16) to calibrate and validate the model for plant VMC and canopy VWC estimation separately with the NDWI. The results of validation with in-situ measurements showed an R2 of 0.44 for the corn plant VWC, an R2 of 0.66 for the corn canopy VWC; and an R2 of 0.78 for the soybean plant VWC, an R2 of 0.85 for the soybean canopy VWC, respectively.

Model-based sensor data fusion of quasi-redundant voltage and current measurements in a lithium-ion battery module

In recent years, battery monitoring systems on cell level have been introduced. These architectures frequently employ current and voltage sensors at each battery cell. In this work, a novel model-based sensor data fusion method is presented, which combines the cell's individual current and voltage measurements of a battery module in order to reduce the measurement uncertainty of each value. Especially with distinct battery cell parameters, which typically arise with ongoing aging, the algorithm is superior to common approaches like averaging or total current sensing as it can handle uneven current and voltage distribution. A framework is proposed consisting of a weighted least squares regression algorithm in combination with dual Kalman filters for estimating the battery cells' states and parameters simultaneously. The method is validated with experiments and complementary simulations, in which a battery module, consisting of 12 cells, is cycled with specific current profiles. Reference model parameters are determined by supplementary characterization tests. It is shown that the sensor noise is reduced significantly with the presented algorithm, while no systematic error is induced.

Sensor data fusion for responsive high resolution ultrasonic temperature measurement using piezoelectric transducers

Ultrasonic temperature measurement allows for responsive measurements across an entire ultrasonic pathway, unlike most conventional temperature sensors that respond to the temperature at the point of their placement only after a notable response time. The high cost of required ultrasonic instrumentation can be reduced substantially by using ultrasonic oscillating temperature sensors (UOTS) consisting of inexpensive narrowband piezo transducers and driving electronics. An UOTS produces sustained oscillations at a frequency that relates to the temperature of the medium between the transducers. The existence of thermal hysteresis in UOTS readings, observed experimentally and apparently related to the fundamental properties of piezoelectric materials, makes conversion of the output frequency readings to the temperature values ambiguous. This makes it complicated to calibrate and use UOTS on their own. In the reported experiment (heating, then naturally cooling of a water vessel equipped with both UOTS and conventional sensors), this hysteresis was solved by fusing UOTS data with conventional temperature sensor readings. As the result, the combination of one UOTS plus one conventional reference sensor allowed improving both the temperature resolution and responsiveness of the latter and ambiguity of the readings of the former. Data fusion effectively led to calibrating the UOTS at every change of the conventional sensor's reading, removing any concerns related to the thermal expansion/contraction of the ultrasonic pathway itself and/or hysteresis of piezoelectric transducers.

Data Fusion Method and Probabilistic Pairing Approach in Elastic Constants Measurement by Resonance Ultrasound Spectroscopy

Resonant ultrasound spectroscopy (RUS) is a nondestructive technique for measuring the stiffness parameters of solid materials. Trying to achieve an improved automatic, elastic coefficients measurement via RUS, this work proposes a data fusion method based on Bayesian formulation and a probabilistic pairing approach. Several sets of RUS data have been acquired by measuring a novel piezoelectric crystal Ca3NbAl0.5Ga0.5Si2O14 (CNAGS) with a self-designed RUS measurement system. In order to suppress the influence of the phenomenon of frequency shifting, missing, and overlapping, a repeated measurement is executed to achieve the multiple measurement data fusion by using Bayesian formulation, and the posterior distribution of each parameter is estimated by reversible jump Markov chain Monte Carlo (RJ-MCMC) algorithm. Besides, an automatic probabilistic pairing method is proposed to solve the problem of frequency crossing so as to facilitate the frequency pairing process. The RUS measuring apparatus, the algorithm principle, and the calculation process are illustrated systematically in this paper. The algorithm is tested and analyzed based on the measured RUS data, and full sets of elastic coefficients of CNAGS specimen has been obtained in this work. Compared to the IEEE standard resonant method, the method proposed in this work can accomplish the characterization within a shorter time, and the relative error of the results is less than 2%.

Multi-rate data fusion for dynamic displacement measurement of beam-like supertall structures using acceleration and strain sensors

Accurate measurement of dynamic displacement is important for the structural health monitoring and safety assessment of supertall structures. However, the displacement of a supertall structure is difficult to be accurately measured using the conventional methods because they are either inaccurate or inconvenient to be set up in practice. This study provides an accurate and economical method to measure dynamic displacement of supertall structures accurately by fusing acceleration and strain data, which are generally available in the structural health monitoring system. Dynamic displacement is first derived from the measured longitudinal strains based on geometric deformation without requiring mode shapes. An optimization technique is utilized to optimize the deployment of strain sensors for achieving more accurate strain-derived displacement. The strain-derived displacement is then combined with measured acceleration via a multi-rate Kalman filtering approach. Applications to a numerical supertall structure and a laboratory cantilever beam verify that the proposed method accurately estimates displacement including both high-frequency and pseudo-static components, under different noise cases and sampling rates. A full-scale field test on the 600 m-high Canton Tower is implemented to validate the applicability of the proposed method to real supertall structures. Error analysis demonstrates that the data fusion displacement is more accurate than the global position system-measured displacement in the time and frequency domains.

Resolving Three-Dimensional Surface Motion with InSAR: Constraints from Multi-Geometry Data Fusion

Interferometric synthetic aperture radar (InSAR) technology has been widely applied to measure Earth surface motions related to natural and anthropogenic crustal deformation phenomena. With the widespread uptake of data captured by the European Space Agency’s Sentinel-1 mission and other recently launched or planned space-borne SAR missions, the usage of the InSAR technique to detect and monitor Earth surface displacements will increase even more in the coming years. However, InSAR can only measure a one-dimensional motion along the radar line of sight (LOS), which makes interpretation and communication of InSAR measurements challenging, and can add ambiguity to the modelling process. Within this paper, we investigate the implications of the InSAR LOS geometry using simulated and observed deformation phenomena and describe a methodology for multi-geometry data fusion of LOS InSAR measurements from many viewing geometries. We find that projecting LOS measurements to the vertical direction using the incidence angle of the satellite sensor (and implicitly assuming no horizontal motions are present) may result in large errors depending on the magnitude of horizontal motion and on the steepness of the incidence angle. We quantify these errors as the maximum expected error from simulated LOS observations based on a Mogi deformation model. However, we recommend to use LOS observations from several image geometries wherever data are available, in order to solve for vertical and E–W oriented horizontal motion. For an anthropogenic deformation phenomenon observed in seven independent InSAR analyses of Envisat SAR data from the Sydney region, Australia, we find that the strong horizontal motion present could lead to misinterpretation of the actual motion direction when projecting LOS measurements to vertical (uplift instead of subsidence). In this example, the difference between multi-geometry data fusion and vertical projection of LOS measurements (at an incidence angle of 33.8°) reach up to 67% of the maximum vertical displacement rate. Furthermore, the position of maximum vertical motion is displaced horizontally by several hundred metres when the LOS measurements are projected.

Measurement Data Fusion Based on Optimized Weighted Least-Squares Algorithm for Multi-Target Tracking

The outliers remove, the classification of effective measurements, and the weighted optimization method of the corresponding measurement are the main factors that affect the positioning accuracy based on range-based multi-target tracking in wireless sensor networks. In this paper, we develop an improved weighted least-square algorithm based on an enhanced non-naive Bayesian classifier (ENNBC) method. According to the ENNBC method, the outliers in the measurement data are removed effectively, dataset density peaks are found quickly, and remaining effective measurements are accurately classified. The ENNBC method improves the traditional direct classification method and took the dependence among continuous density attributes into account. Four common indexes of classifiers are used to evaluate the performance of the nine methods, i.e., the normal naive Bayesian, flexible naive Bayesian (FNB), the homologous model of FNB (FNB ROT ), support vector machine, k-means, fuzzy c-means (FCM), possibilistic c-means, possibilistic FCM, and our proposed ENNBC. The evaluation results show that ENNBC has the best performance based on the four indexes. Meanwhile, the multi-target tracking experimental results show that the proposed algorithm can reduce the root-mean-squared error of the position compared with the extended Kalman filter. In addition, the proposed algorithm has better robustness against large localization and tracking errors.

Data fusion based multi-rate Kalman filtering with unknown input for on-line estimation of dynamic displacements

Structural dynamic displacement is one of the most important measurands that describe the dynamic characteristics of a structure. However, accurate measurement of dynamic displacements of a civil infrastructure is still a challenging task. To solve the difficulties and drawbacks of direct dynamic displacement measurement, the approach of multi-rate Kalman filtering for the data fusion of displacement and acceleration measurement was developed. Recently, an improved technique for dynamic displacement estimation by fusing biased high-sampling rate acceleration and low-sampling rate displacement measurements has been proposed. However, this technique can only take constant acceleration bias into account. In this paper, based on the algorithm of Kalman filter with unknown input recently developed by the authors, dynamic displacement is on line estimated based on multi-rate data fusion of high-sampling rate acceleration with time-varying bias and low-sampling rate displacement measurements. The time history of time-varying acceleration bias is treated as “unknown input” in the algorithm of Kalman filter with unknown input to overcome the limitations of the previous technique. Some numerical examples with linear or polynomial acceleration bias are used to demonstrate the effectiveness of the proposed approach for on line estimation of structural dynamic displacement.

Comparison of three least-squares methods for fusion of data from radar and depth sensors applied for persons’ monitoring

The research reported in this paper is related to the fusion of measurement data from the impulse-radar sensors and infrared depth sensors applied in a system for unobtrusive monitoring of elderly persons. Three least-squares methods – ordinary least squares, total least squares and weighted total least squares – used for fusion of measurement data are compared with respect to their capacity of decreasing the uncertainty of position estimation in a series of experiments which involved the tracking of a moving person.

Lunar landing navigation supported by ranging to an orbiter

Precise and safe landings on celestial bodies as the Moon are needed for the further exploration and potential future exploitation of solar system bodies. The current navigation systems for landing probes are based on sensor data fusion of inertial and optical sensor measurements to determine the current flight state with respect to the target body and the desired landing site. To improve the accuracy of lunar landing navigation, radiometric measurements (range and range-rate) with respect to an orbiting vehicle have been proposed in several studies. This paper analyzes the impact of range and range-rate measurements from an orbiting vehicle on the navigation accuracy. For that purpose, a baseline configuration of an optical navigation system is used which has already been taken as a reference for investigating the effect of ground-based beacons. First, the influence of the orbiter trajectory in relation to the landing trajectory is investigated. In particular, the line of sight between orbiter and lander permits a preliminary assessment of the navigation performance. Later, the effects of sensor noise for all measurements and of the orbiter ephemeris error are also evaluated. Finally, for a selected scenario, it is demonstrated that radiometric measurements with respect to an orbiter can reduce the positioning error by about 40%.

SVM-based multisensor data fusion for phase concentration measurement in biomass-coal co-combustion.

In this paper, the electrical method combines the electrostatic sensor and capacitance sensor to measure the phase concentration of pulverized coal/biomass/air three-phase flow through data fusion technology. In order to eliminate the effects of flow regimes and improve the accuracy of the phase concentration measurement, the mel frequency cepstrum coefficient features extracted from electrostatic signals are used to train the Continuous Gaussian Mixture Hidden Markov Model (CGHMM) for flow regime identification. Support Vector Machine (SVM) is introduced to establish the concentration information fusion model under identified flow regimes. The CGHMM models and SVM models are transplanted on digital signal processing (DSP) to realize on-line accurate measurement. The DSP flow regime identification time is 1.4 ms, and the concentration predict time is 164 μs, which can fully meet the real-time requirement. The average absolute value of the relative error of the pulverized coal is about 1.5% and that of the biomass is about 2.2%.