This paper investigates the consensus tracking problem of leader-follower multi-agent systems. Different from most existing works, dynamics of all the agents are assumed completely unknown, whereas some input-output d...
This paper investigates the consensus tracking problem of leader-follower multi-agent systems. Different from most existing works, dynamics of all the agents are assumed completely unknown, whereas some input-output data about the agents are available. It is well known from the Willems et al. Fundamental Lemma that when inputs of a linear time-invariant (LTI) system are persistently exciting, all possible trajectories of the system can be represented in terms of a finite set of measured input-output data. Building on this idea, the present paper proposes a purely data-driven distributed consensus control policy which allows all the follower agents to track the leader agent’s trajectory. It is shown that for a linear discrete-time multi-agent system, the corresponding controller can be designed to ensure the global synchronization with local data. Even if the data are corrupted by noises, the proposed approach is still applicable under certain conditions. Numerical examples corroborate the practical merits of the theoretical results.
作者：
Xiaofei ZhangHongbin MaWenchao ZuoMan LuoSchool of Automation
Beijing Institute of TechnologyBeijing 100081 School of Vehicle and Mobility
Tsinghua UniversityBeijing 100084China School of Automation
Beijing Institute of Technologyand also with the State Key Laboratory of Intelligent Control and Decision of Complex Systems(Beijing Institute of Technology)Beijing 100081China School of Automation
Beijing Institute of TechnologyBeijing 100081and he is with Beijing Institute of Electronic System EngineeringBeijing 100854China School of Automation
Beijing Institute of TechnologyBeijing 100081and she is with Ant GroupBeijing 310013China
Random vector functional ink(RVFL)networks belong to a class of single hidden layer neural networks in which some parameters are randomly *** network structure in which contains the direct links between inputs and out...
详细信息
Random vector functional ink(RVFL)networks belong to a class of single hidden layer neural networks in which some parameters are randomly *** network structure in which contains the direct links between inputs and outputs is unique,and stability analysis and real-time performance are two difficulties of the controlsystems based on neural *** this paper,combining the advantages of RVFL and the ideas of online sequential extreme learning machine(OS-ELM)and initial-training-free online extreme learning machine(ITFOELM),a novel online learning algorithm which is named as initial-training-free online random vector functional link algo rithm(ITF-ORVFL)is investigated for training *** link vector of RVFL network can be analytically determined based on sequentially arriving data by ITF-ORVFL with a high learning speed,and the stability for nonlinear systems based on this learning algorithm is *** experiment results indicate that the proposed ITF-ORVFL is effective in coping with nonparametric uncertainty.
An event-triggered control (ETC) system transmits data packages and updates control inputs only when the predefined criterion is *** this way,network communication and computing resources are scheduled more reasonably...
详细信息
An event-triggered control (ETC) system transmits data packages and updates control inputs only when the predefined criterion is *** this way,network communication and computing resources are scheduled more reasonably in contrast to the traditional periodic sampling ***-gain approach proposed in recent literatures is a new modeling method to deal with nonlinear ETC *** from traditional ETC models,stability criteria are proposed in the form of input to state stability (ISS) gain to design the triggering *** paper introduces additional dynamic variables in this model and proposes a small-gain based dynamic event-triggered *** conditions to guarantee the stability of the system are derived with the help of cyclic-small-gain theorem and Zeno behaviors are avoided to ensure the feasibility of this method in practical *** simulations are given to demonstrate the effectiveness of our approach.
In this paper,according to the different dynamic models of the support phase and the swing phase,the joint torque compensation of the support phase and the swing phase are given to compensate the joint torque of the h...
详细信息
In this paper,according to the different dynamic models of the support phase and the swing phase,the joint torque compensation of the support phase and the swing phase are given to compensate the joint torque of the hydraulic quadruped robot controlled by the virtual model control(VMC).This compensation can increase the tracking effect of the foot end on the swing trajectory,reduce the elasticity and damping coefficient of the virtual model method,and increase the flexibility of the *** this paper establishes the touchdown detection function based on the joint torque obtained by the Lagrange dynamics equation and the joint torque obtained by the force sensor to determine the touch state of the foot of the quadruped robot to switch the phase of the *** the end of this paper,simulation and experiment prove that the compensation model and touchdown detection function model have feasibility and correctness.
This paper studies the finite-time tracking consensus in probability of multi-agent systems with noises and time delay. By using the method of pinning control and observer, the system both with and without the leader ...
详细信息
An important question in data-driven control is how to obtain an informative dataset. In this work, we consider the problem of effective data acquisition of an unknown linear system with bounded disturbance for both o...
详细信息
This article deals with model- and data-based consensus control of heterogenous leader-following multi-agent systems (MASs) under an event-triggering transmission scheme. A dynamic periodic transmission protocol is de...
详细信息
Temporal Action Detection is a research hotspot in Video *** is necessary not only to give the specific moments of the beginning and end of each action instance in the video,but also to give the category of the action...
详细信息
Temporal Action Detection is a research hotspot in Video *** is necessary not only to give the specific moments of the beginning and end of each action instance in the video,but also to give the category of the action *** present,most of the methods in temporal action detection are divided into two steps:The video is divided into a series of video clips to determine whether each clip is an action instance,and the fragments that are not an action instance are deleted;Segments that may be action instances are then classified to get the final *** present,the difficulties in action detection research mainly include the following two points:1) The boundary is not *** from action recognition,action detection requires precise positioning,but an action in life is often not very *** is also the reason why the mean average precision(mAP) of action detection is low at present.2) The time span is *** life,an action often spans a very long *** movements such as waving can last for a few seconds,while long movements such as rock climbing or cycling can last for tens of minutes,which makes it extremely difficult for us to extract *** view of the above difficulties,this paper proposes a large receptive field boundary matching network(LRFBMN) model which takes advantage of the relationship between proposals to improve the accuracy of proposal *** model is mainly divided into two parts:1) Clipping feature map is processed by large kernel convolution,and then proposal feature is generated by ROI-pooling;2) The proposals are arranged in a certain order to form a fixed graph,and the information exchange between graph nodes is realized by using convolution with *** experiments,this model is 2.35% higher than baseline and 1.06% higher than state-of-the-art in THUMOS14 data set.
One of the challenges of RGB-D-based robotic grasping technology is how to make full use of these two complementary heterogeneous data sources while ensuring real-time performance. The existing robotic grasping method...
详细信息
One of the challenges of RGB-D-based robotic grasping technology is how to make full use of these two complementary heterogeneous data sources while ensuring real-time performance. The existing robotic grasping methods mainly extract information from single-mode images or use a one-branch network to process RGB and depth images. These methods are insufficient in fully fusing the effective information of the two modes, which limits the anti-jamming performance. In this work,we propose Attention Dense Fusion Network（ADFNet）, a novel RGB-D based robotic grasping system that directly predicts the optimal grasping pose from RGB-D images. Our system uses a two-branch network with a heterogeneous framework that processes RGB and depth image information separately and retain the original structure of each data source. After applying dense fusion networks at different scales, the high-dimensional features in the RGB and depth branches are embedded and fused at the pixel level. Besides, we incorporate attention mechanisms to effectively suppress the independent background regions of the feature map and enhance the significant features, which improves the prediction accuracy. We conduct qualitative and quantitative experiments on the standard Cornell Grasp Dataset. The experimental results show that ADFNet can effectively improve the prediction accuracy to 98.9 %, which is better than existing methods while ensuing real-time performance.
In this work,we use a hierarchical architecture based on detector-classifier for gesture recognition *** the operation of the architecture,the detector,which is essentially the switch of the classifier,is always *** t...
详细信息
In this work,we use a hierarchical architecture based on detector-classifier for gesture recognition *** the operation of the architecture,the detector,which is essentially the switch of the classifier,is always *** the output of the detector is true,then the classifier is activated and returns a classification label for the input video *** work focuses on the improvement of detectors and *** the detector,we introduce an attention mechanism to guide the network to focus on the space and channel where the gesture is *** the classifier,based on the RGB information stream,we use an independent branch to extract the features of the depth stream,and finally merge the two *** gestures move in a three-dimensional space,depth information can make up for the lack of RGB *** show that on the Egogesture test set,our detector achieves 98.86% accuracy on RGB input,while the classifier achieves 93.85% *** the same time,our gesture recognition architecture can fully meet the real-time requirements.
暂无评论