Non-cooperative target track prediction method based on Attention-LSTM

Wireless Sensor Network WSN is consisted of nodes with different sizes and a specific goal. Tracking applications are very important in WSNs. This study proposes a method for reducing energy consumption in WSNs, considering target tracking. A prediction method is also presented with the help of MAC Medium Access Control layer in which a few number of nodes are in active mode to track the target while the rest of the nodes are switched into sleeping mode. Energy consumption was reduced by decreasing the number of nodes involved in target tracking and activating merely a limited number of necessary nodes. The conducted experiments suggest that the proposed algorithm has a much better performance compared to similar algorithms in different conditions, including energy consumption in different communication ranges, effect of the number of nodes on energy consumption, the number of nodes involved in tracking in terms of communication range, and losing the target. This means the proposed method is superior and more efficient. When the target enters the simulation area, nodes of the network identify the target node through the proposed method (which is the optimum method) and track it as long as it is inside the simulation area. Performance criteria, including energy, other nodes, and throughput were used in simulation. The results showed that the proposed method has the optimum performance and reduces energy consumption of the network during tracking. The results of the simulation showed that the proposed method reduces energy consumption to an acceptable extent, compared to previous methods, in addition to increasing the accuracy of target tracking. Keywords: Active nodes, MAC layer, Energy efficiency, Target tracking, Wireless sensor networks, Target tracking.

The relative pose (position and attitude) measurement of space noncooperative targets is very important for on-orbit servicing activities, such as target tracking, approaching, and capturing. The traditional methods rarely consider the instability of feature extraction and image blurring caused by target tumbling. In this paper, a method based on the maximum outer contour (MOC) recognition is proposed to measure the pose of the target. Different feature extraction algorithms can simultaneously achieve close- and long-range measurement tasks. First, the trailing image is restored by the image enhancement method. Second, the “rough + fine” combination recognition method is used for contour extraction and connected component labeling of the restored image, and the target feature extraction time is reduced to one-third of traditional methods. Furthermore, the elliptical surface on the MOC is fitted by the least squares method (LSM), and the ellipse parameters (i.e., the center position, the long- and short-axes, and the deflection angle) are extracted. The accuracy of the target recognition is improved. Third, for the close-range measurement, based on the detected ellipse parameter, the pose of the noncooperative target is solved by the binocular imaging algorithm of the space circle; for the long-range measurement, the contour centroid of the target is calculated by the detected MOC, and the position of the target is solved by the LSM. Moreover, the effectiveness of the method is verified by the OpenSceneGraph numerical simulation system. Finally, an experimental system consisting of a binocular camera, a UR5 manipulator, and a satellite mockup was built. The experimental results verified the proposed method.

To ensure the safety of the space station and improve the accuracy of the estimated trajectory tracking of noncooperative target, an optical imaging micro-/nanosatellite based on APSO-SSUKF (adaptive particle swarm optimization-spherical simplex unscented Kalman filter) is proposed to track low-orbit target using angle-only measurement. First, the algorithm considers the effect of J2 perturbation, uses the angle-only data as the observation vector, and uses spherical simplex unscented Kalman filter (SSUKF) to reduce the cost of calculation of the UKF in space noncooperative target tracking. Secondly, it is proposed to use the actual and theoretical covariance of the innovation sequence for real-time estimation of measurement noise, designing the adaptive particle swarm optimization (APSO) algorithm for real-time tracking of the process noise in the SSUKF that improves the accuracy of the filter in angle-only tracking. Finally, the tracking simulation of low-orbit satellite is carried out by using optical imaging micro-/nanosatellite, and the result shows that, compared with UKF, SSUKF, and PSO-SSUKF, APSO-SSUKF reduces the root mean square of the error in predicting the position in space target tracking by 45.44%, 35.26%, and 20.94%, and APSO-SSUKF reduces the root mean square of the error in velocity by 45.58%, 33.53%, and 16.33%, respectively; in the angle-tracking target, APSO-SSUKF improves the convergence and estimated accuracy of the algorithm in tracking.

Adaptive data processing in modern radar systems is standard approach. The classic kalman filter, extended kalman filter and many other applications are usually used for state estimate and radar target tracking. Signal and data processing technology improvement allows the more effective target tracking in real time. In this paper, we approach the structure and key features of the optimal particle filtering (PF). The optimization criteria for PF are proposed and analyzed. The dependence of mean square error on the number of particles during track prediction by particle filter is discussed. Using this dependence, we were able to determine the minimum required number of particles to correctly predict the trajectory of the target depending on its dynamics (trajectory complexity). The results of off-line real radar data processing are analyzed.

This paper presents a new methodology for improving the tracking of multiple targets in complex scenes. The new method,Motion Parameter Sharing, incorporates social motion information into tracking predictions. This is achieved by allowing a tracker to share motion estimates within groups of targets which have previously been moving in a coordinated fashion. The method is intuitive and, as well as aiding the prediction estimates, allows the implicit formation of 'social groups' of targets as a side effect of the process. The underlying reasoning and method are presented, as well as a description of how the method fits into the framework of a typical Bayesian tracking system. This is followed by some preliminary results which suggest the method is more accurate and robust than algorithms which do not incorporate the social information available in multiple target scenarios.

Global fixed-time prescribed performance control with fixed-time distributed state observer for non-cooperative target tracking using multiple unmanned surface vehicles.

As the Automatic Dependent Surveillance-Broadcast (ADS-B) system has gained wide acceptance, additional exploitations of the radioed satellite-based information are topics of current interest. One such opportunity includes the augmentation of the communication ADS-B signal with a random biphase modulation for concurrent use as a radar signal. This paper addresses the formulation and analysis of a suitable noncooperative multitarget tracking method for the ADS-B radar system using radar ranging techniques and particle filter algorithms. In addition, the low-update-rate measurement due to the ADS-B system specification is discussed in order to provide acceptable estimation results. Simulation results show satisfactory tracking capability up to several kilometers with acceptable accuracy.

In recent years, MV and kV electronic flat panel imaging devices have become commonplace equipment installed on the gantry of many newer medical LINACs. Such imaging devices are now routinely used in the clinic for verification of correct initial patient setup before start of a treatment fraction. This is done through MV portal images, 2D kV fluoroscopic images, or cone‐beam CT. However, these amorphous‐silicon detectors have not been used to their full potential. With high resolution imaging capture rates of ∼ 10 frames per second, the role of these MV and kV detectors can be additionally expanded to include real‐time target tracking during the actual radiation treatment delivery process.This lecture will provide an overview of recent research in the use of MV and kV imagers for real‐time tracking purposes. In addition, an explanation of the basic theory and methods necessary to perform real‐time tracking will be discussed.Learning Objectives:1. Fundamentals of MV and kV flat panel detectors.2. Image processing and prediction methods used for real‐time MV and kV intheclinic.3. Issues involved in implementing hybrid MV and kV in the clinic.

In order to achieve non-cooperative target tracking and ranging in conditions of a weak echo signal, this paper presents a real-time acquisition, pointing, tracking (APT), and ranging (APTR) lidar system based on single photon detection. With this system, an active target APT mechanism based on a single photon detector is proposed. The target tracking and ranging strategy and the simulation of target APT are presented. Experiments in the laboratory show that the system has good performance to achieve the acquisition, pointing and ranging of a static target, and track a dynamic target (angular velocity around 3 mrad/s) under the condition of extremely weak echo signals (a dozen photons). Meanwhile, through further theoretical analysis, it can be proven that the mechanism has stronger tracking and detection ability in long distance. It can achieve the active tracking of the target with a lateral velocity of hundreds of meters per second at about one hundred kilometers distance. This means that it has the ability of fast long-distance non-cooperative target tracking and ranging, only by using a single-point single photon detector.

Tracking and operating non-cooperative targets at close range are key problems in the field of intelligent space robot. In this paper, three non-cooperative target tracking methods including EKF method, SEKF method and Inf-EKF method are investigated. In addition, efficiency, filtering performance and tracking accuracy of three tracking methods are researched by simulations. The results show that compared with EKF method, SEKF method and Inf-EKF method reduce the calculation and improve the efficiency. Moreover, by comparing filtering performance under different noise levels, it is found SEKF method doesn't show rapid convergence rate but has good anti-noise performance. And through tracking trajectories of feature points, it proves that SEKF method can track targets reliably despite high-level noise.

This article investigates the satellite constellation scheduling problem for noncooperative space-object tracking. Different from the traditional Earth observation tasks, space-object tracking tasks require observation of fast moving targets that may have unpredictable trajectories, and estimation of the position of the targets through cooperative observation of multiple satellites. An agent-based scheduling framework is proposed to enable real-time, distributed scheduling of low Earth orbits constellation satellites for tracking noncooperative space targets. In the proposed framework, a nonlinear filtering of the targets’ trajectories is incorporated, and the observation satellites are dynamically allocated according to the status of the satellites and the states of the noncooperative targets. Specifically, event triggers of tracking space targets for activating task handover among satellites are defined and classified into five categories. The event-trigger processing and task response mechanism are embedded in the proposed scheduling framework. Thereafter, autonomously adjusting the observation-satellite sets can be achieved through distributed negotiation within only a few satellites. The effectiveness of the proposed method is validated by simulating scenarios of tracking targets without or with maneuvers.

Motion prediction of space noncooperative target is an important issue for spacecraft on-orbit service. After obtaining high-precision motion prediction results, the chaser can efficiently plan a motion trajectory to approach the target and then capture it. In this paper, a motion prediction method considering J2 perturbation is proposed for a free-floating noncooperative target. The core idea of this method is to identify dynamic parameters of the target, and then realize the motion prediction through a dynamic model of the target that considering J2 perturbation. In the identification of the dynamic parameters, inertia parameters of the target are preliminarily identified first, then a noise-adaptive unscented Kalman filter (UKF) is applied to roughly identify the dynamic parameters, and finally, the identification precision is further improved through optimization. At the end of this paper, the significance of considering J2 perturbation is demonstrated by numerical simulations, and the effectiveness of the proposed method is verified. Simulation results indicate that J2 perturbation has a great influence on the motion prediction of the target and that the proposed method can yield long-time high-precision motion prediction results. Specifically, under the simulation conditions of this paper, the predicted time obtained by the proposed method is longer than 200 s, and the errors of the motion prediction results of the target's attitude and position are less than 2 deg and 0.02 m, respectively.

Radar receives more and more attention as an important means of access to information, and multi-sensor data fusion and track prediction become a new discipline. Compared with single radar, multi-radar system has the advantage of improving reliability of the system, enhancing the system when coverage, etc. However, this technology faces new problems when utilizing information in the complex environment. In this paper, we used the nearest data association algorithm (NNDA) to extract tracks from multifarious radar data and three spline interpolation method to make different measuring data track registration to the unity of time axis. Through the same period of fuzzy track correlation, we realized the same target track of extraction, calculated the relative radar accuracy by using the least square fitting, and applied radar tracking precision to be the integration of the weighted average method through the fusion of track-to-track reference. Finally, based on the knowledge of neural network technology, we used echo state network (ESN) to predict data, as ESN network training algorithm is very good in effectively solving nonlinear dynamic system specifically of uncertainty model. Through simulation test, we concluded that the multi-sensor data fusion and trajectory prediction model precision accuracy is 43.18 m for uniform motion model; for a sudden turn or variable targets, precision accuracy is 122.7m; for complex moving targets, precision accuracy is 165.3m.

Tropical cyclones are among the most powerful and destructive meteorological systems on Earth. In this paper, we propose a novel deep learning model for tropical cyclone track prediction method. Specifically, the track task is regarded as a time series predicting challenge, and then a deep learning framework by a bidirectional gate recurrent unit network (BiGRU) with attention mechanism is developed for track prediction. This proposed model can excavate the effective information of the historical track in a deeper and more accurate way. Data experiments are conducted on tropical cyclone best-track data provided by the Joint Typhoon Warning Center (JTWC) from 1988 to 2017 in the northwestern Pacific Ocean. Results show that our model performs well for tracks of 6, 12, 24, 48, and 72 h in the future. The prediction results show that our proposed combined model is superior to state-of-the-art deep learning models, including a recurrent neural network (RNN), long short-term memory neural network (LSTM), gate recurrent unit network (GRU), and BiGRU without the use of attention mechanism. In comparison with the methods used by the China Meteorological Administration, Japan Meteorological Agency, and the JTWC, our method has obvious advantages in the mid- to long-term track forecasting, especially in the next 72 h.

Aiming at the requirements of ground simulation and technical verification of space non-cooperative target relative measurement tasks, a space non-cooperative target relative measurement ground simulation system with dual-manipulator cooperative motion is built to realize the accurate simulation of the relative position and attitude between target aircraft and tracking aircraft during close-range spiral approximation process. The system is calibrated by laser tracker and checkerboard target which can establish the conversion relationship between active and passive manipulator. By compensating the calibration results into the system, the relative position and attitude between the target aircraft and tracking aircraft can be accurately obtained as the benchmark information for testing and evaluation and the experimental test and error evaluation of the real-time pose estimation of typical space non-cooperative targets are completed. The experimental test and error evaluation of the typical space non cooperative targets real-time pose estimation are completed based on the system.KeywordsDual-manipulator cooperative motionSystem calibrationNon-cooperative targetPose estimation