This article considers the robust consensus problem of the general linear multiagent system (MAS) subject to both heterogeneous additive stable disturbances and input saturation. Distributed low gain feedback-based dy...
详细信息
This article considers the robust consensus problem of the general linear multiagent system (MAS) subject to both heterogeneous additive stable disturbances and input saturation. Distributed low gain feedback-based dynamic output feedback control protocols are proposed, which do not need the controller interaction. Algebraic Riccati equation and unified $H_{infinity}$ controller design method are employed to design the output feedback control protocols. It is established that under the assumption that each agent is asymptotically null controllable with bounded controls, semiglobal robust consensus can always be reached under the proposed controller. Furthermore, the design method specialized for leader-following consensus is addressed, under the assumption that the Laplacian matrix is diagonalizable, one can design the control protocol only with the number of follower agents. Finally, several simulations are provided to show the effectiveness of our results.
Energy management in the smart home can help reduce residential energy costs by scheduling various energy consumption activities. However, accurately modeling factors, such as user behavior, renewable power generation...
详细信息
Energy management in the smart home can help reduce residential energy costs by scheduling various energy consumption activities. However, accurately modeling factors, such as user behavior, renewable power generation, weather conditions, and real-time electricity prices can be challenging, making the design of an efficient energy management strategy difficult. This article proposes a real-time energy management algorithm based on deep reinforcement learning (DRL) for smart homes equipped with rooftop photovoltaics, energy storage systems, and smart appliances. The algorithm aims to minimize the energy cost while ensuring user comfort. A policy network that can output both discrete and continuous actions is designed to generate actions for different types of devices in a smart home. The proposed DRL-agent is trained using a proximal policy optimization approach with historical data and is used for real-time scheduling. Finally, simulations based on real-world data demonstrate the effectiveness and robustness of the proposed algorithm.
This paper first proposes and solves weakly supervised 3D human pose estimation (HPE) problem in point cloud, via propagating the pose prior within unlabelled RGB-point cloud sequence to 3D domain. Our approach termed...
详细信息
ISBN:
(纸本)9783031200649;9783031200656
This paper first proposes and solves weakly supervised 3D human pose estimation (HPE) problem in point cloud, via propagating the pose prior within unlabelled RGB-point cloud sequence to 3D domain. Our approach termed C3P does not require any labor-consuming 3D keypoint annotation for training. To this end, we propose to transfer 2D HPE annotation information within the existing large-scale RGB datasets (e.g., MS COCO) to 3D task, using unlabelled RGB-point cloud sequence easy to acquire for linking 2D and 3D domains. The self-supervised 3D HPE clues within point cloud sequence are also exploited, concerning spatial-temporal constraints on human body symmetry, skeleton length and joints' motion. And, a refined point set network structure for weakly supervised 3D HPE is proposed in encoder-decoder manner. The experiments on CMU Panoptic and ITOP datasets demonstrate that, our method can achieve the comparable results to the 3D fully supervised state-of-the-art counterparts. When large-scale unlabelled data (e.g., NTU RGB+D 60) is used, our approach can even outperform them under the more challenging cross-setup test setting.
In recent years, the rapid advancement of automation control and intelligent sensing technologies has positioned autonomous driving as a focal point of interest for both academia and industry. As core equipment in mod...
详细信息
This study focuses on the observability of second-order linear time invariant (LTI) systems with incommensurable output matrices through a matrix-weighted graph. Here, the observability of such systems refers to that ...
详细信息
This study focuses on the observability of second-order linear time invariant (LTI) systems with incommensurable output matrices through a matrix-weighted graph. Here, the observability of such systems refers to that the relative outputs have synchronized solutions for the identical LTI systems. Compared with most of existing results, relying on scalar networks (i.e., the weight of edges is a constant), this study investigates the observability in a matrix-weight-based network. Some necessary and sufficient conditions for the observability have been obtained by the space analysis, spectral analysis and matrix decomposition, respectively. Moreover, the relationship between the observability and the connectivity of its interconnection graph is also discussed. Examples and simulations are shown to verify the theoretical results. (c) 2023 The Franklin Institute. Published by Elsevier Inc. All rights reserved.
In order to address the issues of real-time performance and the low dependency between feature channels in fabric defect detection networks, this paper proposes the ESE_YOLOv5 network based on YOLOv5. Firstly, to addr...
In order to address the issues of real-time performance and the low dependency between feature channels in fabric defect detection networks, this paper proposes the ESE_YOLOv5 network based on YOLOv5. Firstly, to address the relative redundancy of the neck detection network feature channels, a relatively lightweight and efficient convolution module is adopted to ensure accuracy while reducing computation and parameter volume. Furthermore, the Efficient Squeeze-Excitation (ESE) module is introduced into the backbone to optimize the dependency of feature channels, which enhances the model's feature extraction capacity and improves detection accuracy. Experimental results show that compared to YOLOv5, the proposed ESE_YOLOv5 model reduces computation and parameter volume while improving accuracy, meeting the needs of fabric defect detection for recognizing fabric defects that have similar characteristics to the background while maintaining real-time performance.
The piezoelectric actuator is one kind of device that can drive nanoscale motion. However, the nonlinear hysteresis effect induced by its natural material greatly degrades its positioning accuracy. To handle this chal...
The piezoelectric actuator is one kind of device that can drive nanoscale motion. However, the nonlinear hysteresis effect induced by its natural material greatly degrades its positioning accuracy. To handle this challenging issue, this work develops a Koopman model predict control (Koopman- MPC) framework for the piezoelectric actuator. Specifically, the Koopman operator theory is adapted for modeling the piezoelectric actuator dynamics. A simple yet powerful linear model spanned in a high-dimensional space is thus constructed to characterize the hysteresis dynamics. Subsequently, upon the established Koopman model, an MPC scheme is put forward for tracking control of piezoelectric actuators. Therein, by sustained optimizing a cost function containing future outputs and control increments, the control input is obtained. Moreover, extensive tracking simulations are carried out on a simulated piezoelectric actuator for verifying the feasibility and effectiveness of the Koopman- Mpc scheme.
Deep perception of the unmanned surface vehicle's surroundings is an inaccessible part of its fully autonomous navigation mission. The existing methods, whether based on traditional stereo matching or deep learnin...
Deep perception of the unmanned surface vehicle's surroundings is an inaccessible part of its fully autonomous navigation mission. The existing methods, whether based on traditional stereo matching or deep learning, do not fully consider the characteristics of water environment, resulting in severe error depths in weak textures (sky, calm lake) and water reflections regions, that increases the risk of running aground or collision. What is worse that there is not a public dataset for depth estimation in the water environment. Therefore, this work proposes a self-supervised model for depth estimation named Water Depth Perception Network (WDNet) to address these problems. The decoder of this network has a wider receptive field and can effectively handle the depth error in the weak texture region. Besides, the WDNet is trained with a novel and effective loss function which assist the network to reduce errors in sky and water region, and some indexes are proposed to evaluate the model's performances in sky and water region. Finally, our proposed WDNet achieves a 0.1056 absolute relative error in ranging, the average number of error pixels in the sky area drops from 15803.87 to 580.91, which only accounted for 0.29% of the image, and the error in water region drops from 51.04 to 6.75, all of them are superior to the performance of baseline model.
Based on the fact that the traditional probability distribution entropy describing a local feature of the system cannot effectively capture the global topology variations of the network, some indicators constructed by...
详细信息
Based on the fact that the traditional probability distribution entropy describing a local feature of the system cannot effectively capture the global topology variations of the network, some indicators constructed by the network adjacency matrix and Laplacian matrix come into being. Specifically, these measures are based on the eigenvalues of the scaled Laplace matrix, the eigenvalues of the network communicability matrix, and the spectral entropy based on information diffusion that has been proposed recently, respectively. In this article, we systematically study the dependence of these measures on the topological structure of the network. We prove from various aspects that spectral entropy has a better ability to identify the global topology than the traditional distribution entropy. Furthermore, the indicator based on the eigenvalues of the network communicability matrix achieves good results in some aspects while, overall, the spectral entropy is able to identify network topology variations from a global perspective.
Different from fine-tuning models pre-trained on a large-scale dataset of preset classes, class-incremental learning (CIL) aims to recognize novel classes over time without forgetting pre-trained classes. However, a g...
详细信息
ISBN:
(数字)9783031198212
ISBN:
(纸本)9783031198205;9783031198212
Different from fine-tuning models pre-trained on a large-scale dataset of preset classes, class-incremental learning (CIL) aims to recognize novel classes over time without forgetting pre-trained classes. However, a given model will be challenged by test images with finer-grained classes, e.g., a basenji is at most recognized as a dog. Such images form a new training set (i.e., support set) so that the incremental model is hoped to recognize a basenji (i.e., query) as a basenji next time. This paper formulates such a hybrid natural problem of coarse-to-fine few-shot (C2FS) recognition as a CIL problem named C2FSCIL, and proposes a simple, effective, and theoretically-sound strategy Knowe: to learn, freeze, and normalize a classifier's weights from fine labels, once learning an embedding space contrastively from coarse labels. Besides, as CIL aims at a stability-plasticity balance, new overall performance metrics are proposed. In hat sense, on CIFAR-100, BREEDS, and tieredimageNet, Knowe outperforms all recent relevant CIL or FSCIL methods.
暂无评论