Abstract ：

Anticipating the future behavior of pedestrians is a crucial part of deploying Automated Driving Systems (ADS) in urban traffic scenarios. Most recent works utilize a convolutional neural network (CNN) to extract visual information, which is then input to a recurrent neural network (RNN) along with pedestrian-specific features like location and speed to obtain temporal features. However, the majority of these approaches lack the ability to parse the relationships of the related objects in the specific traffic scene, which leads to omitting the interactions between the pedestrians and the interactions between the pedestrians and the traffic. For this purpose, we propose a graph-structured model which can dig out pedestrians' dynamic constraints by constructing a traffic-aware scene graph within each frame. In addition, to capture pedestrian movement more effectively, we also introduce a temporal feature representation model, which first uses inter-frame and intra-frame GRU (II-GRU) to mine inter-frame information and intra-frame information together, and then employs a novel attention mechanism to adaptively generate attention weights. Extensive experiments on the JAAD and PIE datasets prove that our proposed model is effective in reaching and enhancing the state-of-the-art performance. © 2022 IEEE.

Cite：

Copy from the list or Export to your reference management。

Abstract ：

This article proposes a new terminal cooperative guidance law to control both impact time and impact angle in three-dimensional engagement. The presented guidance law employs on and normal line-of-sight accelerations to control the impact time and angle together. Two parts are included in this guidance law, that is, the guidance laws with impact time constraint and impact angle constraint. The first is designed based on fixed-time consensus theory to fulfill the salvo attack of the target, and the second is developed based on fixed-time nonsingular fast terminal sliding mode control method to achieve the desired impact angles. The stability analysis of this guidance law based on Lyapunov theory is also presented in detail. Moreover, the maneuvering target can be intercepted successfully with the presented guidance law, and the fixed-time stability and less control energy are well ensured. Additionally, the effectiveness and applicability of the proposed guidance scheme are explicitly verified through simulation tests.

Keyword ：

Cooperative guidance law fixed-time consensus theory fixed-time stability impact time and angle maneuvering target

Cite：

Copy from the list or Export to your reference management。

Abstract ：

For the problem of multiple missiles cooperatively intercepting a maneuvering target, based on finite-time consensus theory, a three-dimensional (3D) cooperative guidance law with line-of-sight (LOS) angle constraints is proposed. Firstly, based on 3D missiles-target relative motion equations, the multi-missile cooperative guidance model with the LOS angle constraint is constructed. Then, the process of cooperative guidance law design is divided into two stages. In the first stage, the acceleration command on the LOS direction is designed based on the finite-time consensus theory, which can guarantee the consensus of all missiles' impact times in finite time. In the second stage, a 3D impact-angle-control guidance law is proposed based on the approaches of nonsingular terminal sliding mode control and fixed-time convergence theory to intercept a maneuvering target, with the use of the approaches of nonsingular terminal sliding mode control and fixed-time convergence theory, the acceleration command on the direction of perpendicular to the LOS is developed, which can ensure that all the LOS angles converge to the desired terminal LOS angle in fixed-time. An extended state observer is constructed to estimate the target acceleration. Finally, numerical simulations of three missiles intercepting a maneuvering target demonstrate the effectiveness of the proposed guidance law. Key words: © 2022, Editorial Department of Control Theory & Applications. All right reserved.

Keyword ：

Electronic guidance systems Equations of motion Guided missiles Sliding mode control State estimation

Cite：

Copy from the list or Export to your reference management。

Abstract ：

3D human action recognition has received increasing attention due to its potential application in video surveillance equipment. To guarantee satisfactory performance, previous studies are mainly based on supervised methods, which have to add a large amount of manual annotation costs. In addition, general deep networks for video sequences suffer from heavy computational costs, thus cannot satisfy the basic requirement of embedded systems. In this paper, a novel Motion Guided Attention Learning (MG-AL) framework is proposed, which formulates the action representation learning as a self-supervised motion attention prediction problem. Specifically, MG-AL is a lightweight network. A set of simple motion priors (e.g., intra-joint variance, inter-frame deviation, intra-joint variance, and cross-joint covariance), which minimizes additional parameters and computational overhead, is regarded as a supervisory signal to guide the attention generation. The encoder is trained via predicting multiple self-attention tasks to capture action-specific feature representations. Extensive evaluations are performed on three challenging benchmark datasets (NTU-RGB+D 60, NTU-RGB+D 120 and NW-UCLA). The proposed method achieves superior performance compared to state-of-the-art methods, while having a very low computational cost. © 2022 IEEE.

Keyword ：

3D human action recognition; motion attention; prior knowledge; self-supervised learning

Cite：

Copy from the list or Export to your reference management。

Abstract ：

A nonsingular terminal sliding mode (NTSM) control scheme with fixed-time convergence is proposed for a class of second-order nonlinear systems subject to matched uncertainties and external disturbances. A compound sliding surface is firstly constructed to eliminate the singularity and the upper bound of convergence time is estimated. A fixed-time NTSM controller is then developed and a guaranteed upper bound of the closed-loop settling time independent of initial states is presented based on the phase plane analysis and Lyapunov stability theory. Finally, a lot of simulations for a single inverted pendulum system are conducted to validate the effectiveness of the proposed methodology. © 2021 Technical Committee on Control Theory, Chinese Association of Automation.

Keyword ：

Nonlinear systems Pendulums Sliding mode control

Cite：

Copy from the list or Export to your reference management。

Cite：

Copy from the list or Export to your reference management。

Abstract ：

This article proposes a new terminal cooperative guidance law with impact time constraint in three-dimensional (3D) engagement. Two parts are comprised by this guidance scheme to control the impact time and fulfill the interception. The guidance law along the line-of-sight (LOS) direction is first designed based on finite time consensus protocol to share time-to-go values among missiles and reach the consensus. Meanwhile, the guidance law on the LOS normal direction is developed based on the fast finite time control method to achieve the interception. The stability analysis of the proposed guidance law based on the Lyapunov theory is also demonstrated in detail. Moreover, the maneuvering target can be intercepted successfully under the presented control algorithm, and the guidance system can fulfill stability within finite time. Additionally, the effectiveness and applicability of the proposed guidance scheme are explicitly verified through simulation tests.

Keyword ：

Cooperative guidance law fast finite time control method finite time consensus protocol impact time constraint maneuvering target

Cite：

Copy from the list or Export to your reference management。

Abstract ：

三维视觉技术是智能机器人的一个热点研究方向,它是智能机器人进行环境感知和实现复杂任务的基础。本文围绕智能制造环境,介绍了机器人三维视觉技术的国内外研究现状及热点研究问题,阐述了它目前在智能制造中的应用,最后对机器人三维视觉技术的未来研究方向进行了探讨。

Keyword ：

三维视觉 智能机器人 智能制造

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Pedestrian trajectory prediction is an important but difficult task in self-driving or autonomous mobile robot field because there are complex unpredictable human-human interactions in crowded scenarios. There have been a large number of studies that attempt to understand humans’ social behavior. However, most of these studies extract location features from previous one time step while neglecting the vital velocity features. In order to address this issue, we propose a novel feature-cascaded framework for long short-term network (CF-LSTM) without extra artificial settings or social rules. In this framework, feature information from previous two time steps are firstly extracted and then integrated as a cascaded feature to LSTM, which is able to capture the previous location information and dynamic velocity information, simultaneously. In addition, this scene-agnostic cascaded feature is the external manifestation of complex human-human interactions, which can also effectively capture dynamic interaction information in different scenes without any other pedestrians’ information. Experiments on public benchmark datasets indicate that our model achieves better performance than the state-of-the-art methods and this feature-cascaded framework has the ability to implicitly learn human-human interactions. Copyright c 2020, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.

Cite：

Copy from the list or Export to your reference management。

Abstract ：

The segmentation of sensor logs based on linear analysis is presented in this paper to know the movement of the vehicle on the weighing platforms, which is a new form of scale at first. A line segment is used to represent points where values are not much different. According to the slopes of the different line segments, the sensor logs are segmented. A maximum-likelihood procedure is also proposed to increase the accuracy of the sensor log segmentation. This algorithm is based on Kalman smoothing and single most likelihood replacement (SMLR). The performance of the algorithm is shown by the experiment that the vehicle weigh -in-motion system calculates the weight of the vehicle with a high accuracy of 99.66%. Moreover, the development board can display the result as soon as the weight is calculated. © 2020 IEEE.

Keyword ：

Maximum likelihood Scales (weighing instruments) Vehicles Weigh-in-motion (WIM)

Cite：

Copy from the list or Export to your reference management。