A method of multirate sensor fusion for target tracking and localization using extended Kalman Filter

In this paper, we investigate the problem of setpoint stabilization and station keeping for Remotely Operated Underwater Vehicles. Manual control of such vehicles is a difficult task, due to environmental disturbances and complexity of the dynamics. The development of automatic control schemes to achieve station keeping and setpoint stabilization is complicated by the lack of precise information regarding the position, velocity and attitude of the vehicle with respect to an inertial reference. We propose a methodology for robust control design of a specific ROV model, which combines sensor information from sources of heterogeneous kind, namely a monocular CDD camera, a yaw compass, a depth sensor, and an inertial platform. The main feature of our approach is the incorporation in the controller structure of a suitable model of the disturbances, which reduces the tmcertainty in the vehicle attitude and position determination. The proposed approach is applied to the problem of robust station keeping of a specific model of a small commercial ROV, employed for pipeline inspection and maintenance of submerged offshore structure. Simulations performed on a 4 degrees-of-freedom model of the specific ROV show that the method is well suited to cope with imprecise sensor information and with environmental disturbances due to underwater current effects. Introduction A crucial problem in underwater automation is given by the scarcity of proper sensor information, especially regarding the real-time position of the vehicle with respect to a reference frame. A number of heterogeneous pieces of information (available at different sampling rates) must be made available to the control authority. As an example, typical information on the vehicle position come from CCD cameras, sonar and ultrasonic transponders, while information of the vehicle attitude are generally retrieved from compasses and rate gyros. In an underwater environment, the performance of these sensors, in terms of accuracy and bandwith, may be quite poor: for instance, signals available from sonar are available only at a very slow sampling rate. Signals from CCD cameras have a narrow bandwidth, and should not be differentiated to retrieve the vehicle velocity to avoid delays and high levels of noise in the control loop. On the other hand, signals from rate gyros and inertial platforms must be integrated to recover vehicle velocity and angular velocity, but this necessarily induces offsets and drifts in the measurements. The goal of this paper is to employ methodologies from nonlinear regulation theory to obtain sensor fusion and achieve setpoint control for a general model of small commercial ROVs. Our approach is to employ internal models of sensor drifts and disturbances to obtain sensor fusion directly at the controller level, rather than at the coordination level (that is, the guidance system) of the control authority. Moreover, internal model-based control is shown to improve the capability of the control system to reject environmental disturbances such as the underwater current. Besides the methodological contribution, our approach allows to endow the control and guidance scheme for ROVs with enhanced station-keeping capabilities.

Comparisons of six heating source localisation measurement models are conducted where temperature or ultrasonic time of flight readings provide the measurement update to the extended Kalman filter for estimating the location of a high heat flux spot source on a flat plate. For a particular measurement model, one of two processes are used: (1) directly using the measurements as the measurement vector in the extended Kalman filter or (2) indirectly obtaining the distance from the sensor to the heating source based on the measurement and then using the obtained distance as the measurement vector in the extended Kalman filter. For the direct models, the Jacobian required by the extended Kalman filter is obtained numerically using finite differences from the finite element forward conduction solution. For the indirect models, the derivatives of the distances with respect to the state variables are obtained in closed form. Heating source localisation results and convergence behaviour are compared for the six measurement models investigated. The ellipse from ultrasonic pulse one-way time of flight measurement model produces the best results when considering accuracy of converged solution, ability to converge to the correct solution given different initial guesses, and smoothness of convergence behaviour. Additionally, extended Kalman filter, extended information filter, and least squares inverse methods are compared for a parameter identification to quantify the heat flux and convection coefficient on the plate. All three inversion methods produce similar results which is significant as future work will consider only the extended Kalman filter.

The FORMOSAT-5 (FS-5), the fifth space program initiated in 2008, is currently developed by National Space Organization (NSPO) of the National Applied Research Laboratories (NARL) in Taiwan, Republic of China (ROC). The satellite will be deployed into a sun-synchronous orbit with 720-km altitude and 98.28o inclination angle. It is a threeaxis attitude stabilized spacecraft providing high precision attitude determination and control for a high resolution Remote Sensing Instrument (RSI). To meet the 0.012 deg, 3sigma, per axis, absolute bus attitude knowledge requirement flown-down from RSI for meeting its geo-location accuracy, the micro-Advanced Stellar Compass (ASC) consisting of three Camera Head Units (CHUs) along with four single-axis fiber-optics gyros is mounted on a stable optical bench common to the RSI structure. The ASC is a highly advanced and fully autonomous star tracker produced by the Measurement & Instrument System (MIS) Section of the Orsted Department of the Technical University of Denmark (DTU). The gyroscopes considered in this mission are the fiber-optics rate sensors, -FORS6U, manufactured by Litef, a Northrop Grumman branch in Germany. The -FORS6U was originally designed to meet the requirements of a wide range of air, land, and sea applications, and was later certified by NSPO for space applications after a series of environmental tests. The primary spacecraft attitude and rate determination for the spacecraft Attitude and Orbit Control System (AOCS) will be provided by a gyro-stellar attitude determination system, which utilizes attitude data provided by three CHUs and rate data provided by the four single-axis fiber-optics gyros. To account for the possibility of more than one gyro failures due to lack of space-flight heritage, a gyro-less attitude and rate determination algorithm is developed as a back-up for the AOCS design. This paper describes the detailed development of a gyro-less attitude and rate determination algorithm which uses the computed gyro data (numerical gyro data) instead of physical gyro data as in a typical gyro-stellar attitude determination system. The numerical gyro data are computed using the assumed spacecraft dynamics (or Euler equations of motion). By perturbing the spacecraft kinematic equations, the linearized attitude error equations are obtained. Similarly, by perturbing the spacecraft dynamic equations, the linearized rate error equations are obtained. With the obtained linearized attitude error equations and the linearized rate error equations, a reduced-order (6-state) Extended Kalman Filter (EKF) providing spacecraft attitude and rate estimates is then implemented in the algorithm. The developed gyro-less attitude and rate determination algorithm was incorporated and tested in a 6-DoF nonlinear, high-fidelity simulation model developed for the FS-5 program to assess its performance. Its performance estimates and sensitivities to spacecraft inertia tensor uncertainties, reaction wheel momentum uncertainties, and torque uncertainties produced by torque-rods during normal operations will be presented in the paper.

Performance enhancement of low-cost, high-accuracy, state estimation for vehicle collision prevention system using ANFIS

Mobile robot localization and map building are very important for a mobile robot to perform a navigation-based service task to aid the aged or disabled in an indoor environment (office, facilities or at home). In order to enable mobile robot to perform service tasks autonomously, the information of environment in which the mobile robot moves is needed. To obtain the accurate map, many different sensors system and techniques have been developed. Approaches of mapping can roughly be classified according to the kind of sensor data processed and the matching algorithms [S. Thrun et al., 1996, 1998, 2000, 2004; Chong, K.S et al., 1996; C.-C. Wang et al., 2002; G. Dissanayake et al., 2000; B. barshan et al., 1995; Weimin Shen et al., 2005]. In this paper, we proposed a method of map building using interactive GUI for a mobile robot. The reason we proposed this method is that it is difficult for a mobile robot to generate an accurate map although many kinds of sensors are used. The proposed method enables the operator to modify map built by sensors, compared with the real-time video from web camera by using modification tool in interactive GUI. Laser Range Finder (LRF) was used to get the information of the environment in which mobile robot moving. In order to improve self-localization of mobile robot, Extended Kalman Filter (EKF) was introduced. By the results of simulation and experiments, the developed EKF is effective in self-localization of mobile robot. This paper introduces the architecture of the system and gives some experimental results to verify the effectiveness of the developed system. The rest of the paper consists of 5 sections. Section 2 describes the structure of the proposed system. Section 3 presents the developed algorithm of self-localization of mobile robot and gives some simulation and experimental results to verify the effectiveness of the EKF in improving precision of positioning of mobile robot. Section 4 details the developed LRF data processing algorithm. Section 5 introduces the developed interactive GUI. The experimental results are given in Section 6. Section 7 concludes the paper. 13

In underwater navigation, the global positioning system is unavailable; hence, other solutions must be pursued in the development of navigation systems. In this paper, filters for the position, linear velocity, and acceleration of underwater vehicles are derived by combining a long baseline acoustic positioning system with an inertial navigation system. A dynamic model is devised via state augmentation, including the bias of the pseudo-ranges, which accounts for the effect of the unknown offset between the emitters’ and receivers’ clocks. With this technique, the resulting dynamics are linear, in spite of the original nonlinear nature of the problem. The proposed solution, which includes the explicit dynamic estimation of the bias of the pseudo-ranges, is a linear Kalman filter. The observability of the system is assessed, which allows to establish globally exponentially stable error dynamics. The performance of the solution is evaluated with realistic simulation results, considering sensor noise and discrete-time measurements. Finally, the comparison with the extended Kalman filter (EKF), the unscented Kalman filter (UKF), and the Bayesian Cramer–Rao bound is presented, including Monte Carlo simulations. This comparison shows the goodness of the proposed solution, which converges for all initial conditions and exhibits performance comparable to the EKF and the UKF, whereas the EKF and UKF are shown to lack global convergence.

In this paper, an implementation of an extended target tracking filter using measurements from high-resolution automotive Radio Detection and Ranging (RADAR) is proposed. Our algorithm uses the Cartesian point measurements from the target's contour as well as the Doppler range rate provided by the RADAR to track a target vehicle's position, orientation, and translational and rotational velocities. We also apply a Gaussian Process (GP) to model the vehicle's shape. To cope with the nonlinear measurement equation, we implement an Extended Kalman Filter (EKF) and provide the necessary derivatives for the Doppler measurement. We then evaluate the effectiveness of incorporating the Doppler rate on simulations and on 2 sets of real data.

Localization is an important function for the robots to complete various tasks. For localization, both internal and external sensors are used generally. The odometry is widely used as the method based on the internal sensors, but it suffers from cumulative errors. In the method using the laser range sensor (LRS) which is a kind of external sensor, the estimation accuracy is affected by the number of available measurement data. In our previous study, we applied moving horizon estimation (MHE) to the vehicle localization for integrating the LRS measurement data and the odometry information where the weightings of them are balanced relatively adapting to the number of the available LRS measurement data. In this paper, the effectiveness of the proposed localization method is verified through both numerical simulations and experiments using a 1/10 scale vehicle. The verification is conducted in the situations where the vehicle position cannot be localized uniquely on a certain direction using the LRS measurement data only. We achieve accurate localization even in such a situation by integrating the odometry and LRS based on MHE. We also show the superiority of the method through comparisons with a method using extended Kalman filter (EKF).

In this paper, according to the characteristics of underwater target tracking, extended Kalman filter (EKF) algorithm was applied to address underwater bearing-only non-maneuverable target tracking problem. To ensure the observability in bearing-only passive target tracking, we use two sensor arrays in uniform rectilinear motion to track the single target. EKF is recursive Bayesian filter algorithm based on the linearization of the nonlinearities in the state and the measurement system. And the simulation experiment shows that, to the constant velocity single target, the bearings-only EKF algorithm with dual-array has good results, also the EKF has lower computation complexity than unscented Kalman filter and particle filter. It confirms the effectiveness of the EKF algorithm in solving the underwater bearing-only target tracking problem.

Continuous tracking and detecting position systems are significant services that can using in many applications such as detecting accident cars on Freeways. This paper introduces Artificial intelligent (AI) method to quick discover the accident and determine the position of car on Freeways that will reduce the rate of fatalities. Most tracking and detecting position systems are based on Global Navigation Satellite System (GNSS). But the GNSS signal in certain areas such as buildings and tunnel is unavailable which make a big problem for these systems. To solve this problem, most traditional methods are based on integrated GNSS with other navigation sensors such as accelerometers, gyroscopes, odometers and so on. But these integrated methods still high cost and need to long time processing. Today, most modern Vehicles are equipping with several On-board Car Motion Sensors (OVMS) such as Acceleration Sensors (AS), Wheel Velocity Sensor (WVS), Gravity Sensor (GS), Yaw Rate Sensor (YRS) and additional to GNSS that can improve the safety and also use to determine and detect the position of accident vehicle. This paper develops an approach AI-accident detection system based on integrating GNSS with On-board Vehicle Motion Sensors (OVMS) using Extended Kalman Filter (EKF) algorithm. During GPS outages, the three Acceleration Sensors (3-ASs) are processing to determine velocities and then position of vehicle on three x, y and z axes. While the YRS and GS sensors are using to determine velocities angles on three x, y and z axes, respectively. In same time the three WVS sensors are using to correct velocities errors of three acceleration sensors (3-ASs) until GNSS signal is return again. The proposed AI-accident detection system is tested in different scenarios during GPS outage signal. The results of proposed method can actually detect accident and determine the position of vehicle with high efficiency.

A gyroless deterministic attitude determination algorithm based on simulation of sun in eclipse is stated in this paper and has been compared to stochastic filters like extended Kalman filter and unscented Kalman filter. Attitude determination with low cost sensors such as magnometer and sun sensor results in usage of recursive algorithms such as Kalman filter which has the probability of divergence, but with deterministic point to point algorithm such as the one introduced in this paper we can be sure to have an attitude determination with a fixed maximum error. Proposed method has been compared with Extended Kalman Filter and Unscented Kalman filter due to its modeling error, robustness and amount of calculations. Attitude determination is based on magnetic sensors and sun sensors. Solar panels are modeled for a hexagonal satellite as an alternative for sun sensors and sun vector is derived from panels current.

This paper presents a new scheme for fault detection and diagnosis in spacecraft Attitude Determination (AD) sensors. An integrated attitude determination system, which includes measurements of rate and angular position using rate gyros and vector sensors, is developed. Measurement data from all sensors are fused by a linearized Kalman filter, which is designed based on the system kinematics, to provide attitude estimation and the values of the gyro bias. Using this information the erroneous sensor measurements are corrected, and unbounded sensor measurement errors are avoided. The resulting bias-free data are used in the Fault Detection and Diagnosis (FDD) scheme. The FDD algorithm uses model-based state estimation, combining the information from the rotational dynamics and kinematics of a spacecraft with the sensor measurements to predict the future sensor outputs. Fault isolation is performed through Extended Kalman Filters (EKFs). The innovation sequences of EKFs are monitored by several statistical tests to detect the presence of a failure and to localize the failures in all AD sensors. The isolation procedure is developed in two phases. In the first phase, two EKFs are designed, which use subsets of measurements to provide state estimates and form residuals, which are used to verify the source of the fault. In the second phase of isolation, testing of multiple hypotheses is performed. The generalized likelihood ratio test is utilized to identify the faulty components. In the scheme developed in this paper a relatively small number of hypotheses is used, which results in faster isolation and highly distinguishable fault signatures. An important feature of the developed FDD scheme is that it can provide attitude estimations in presence of fault in one type of sensors.

This paper presents a map-matching-based vehicle localization algorithm for application to automated driving on urban road. Vehicle position estimation of centimeter-level with low-priced commercial sensor setup is one of the key issue in urban automated driving. The information fusion method of localization algorithm utilizes vehicle chassis sensor and Around View Monitoring (AVM) module with four fish-eyed cameras. The proposed localization algorithm consists of three sections: a lane detection, a position correction, and a localization filter. A lane information is extracted from AVM image around the vehicle. This lane information is possible to correct vehicle position by the iterative closest point (ICP) algorithm which estimates the rigid transformation between the lane map and lanes obtained by AVM in real-time. The corrected vehicle position by this transformation is fused with the information of vehicle sensors based on an extended Kalman filter (EKF). In order to achieve higher accuracy, the covariance of the ICP algorithm is estimated by using Haralick's method. The performance of proposed localization is verified through vehicle experiments on proving ground and actual urban road.

This paper discusses the attitude determination of the CNUSAIL-1 cube-satellite. The primary mission of the CNUSAIL-1 is sail deployment and operation in low Earth orbit, and the secondary mission is to look into influence of the sail deployment on satellite attitude and orbit. The attitude determination strategy is proposed depending on three mission phases, and its performance and applicability are verified through numerical simulations. This study considers the following sensors: Sun sensors and a three-axis magnetometer as attitude reference sensors, and a three-axis MEMS gyroscope as an inertial attitude sensor. Because sensors used for cube satellites have relatively low performances and worse noise characteristics, an Extended Kalman filter (EKF) is applied to attitude determination. Additionally, it has the merits to deal with the Gaussian noises and to predict the attitude even with no measurements from reference attitude sensors, especially in the eclipse of the cube satellite. The performance of the EKF is compared to a deterministic attitude determination technique, QUEST(QUaternion ESTimation).

This paper presents two types of extended Kalman filter (EKF) and two types of unscented Kalman filter (UKF) based on vertical railway vehicle models for parameters estimation of secondary suspensions. Due to track irregularities, the random vertical velocity of the track can be approximated as a zero-mean Gaussian white noise and it is used to excite the dynamic model of the railway vehicle. Under this approximation, the variance of the vertical velocity of the track, which is affected by the track roughness level and vehicle velocity, can introduce uncertainty into the system. Based on the random track irregularity, two cases are proposed to determine how the track irregularities enter the system. One case uses the vertical velocity and displacement of the track as inputs of the system and assumes that the state variables are corrupted by the Gaussian noises. The other case assumes that the vertical velocity of the track is the process noise of the system. Based on these two cases, two types of EKF and UKF are developed to estimate the parameters of the secondary suspensions. In order to study the performances of the proposed EKFs and UKFs, several simulation experiments using linear and nonlinear model are carried out that consider the uncertainties of the random track.