检索结果-内蒙古大学图书馆

International Conference on control, Automation, Robotics and Vision (ICARCV)

作者： Cui Wei Ye-Hwa Chen Yanling Wei Institute of Intelligent Manufacturing Nanjing Tech University Nanjing China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China George W. Woodruff School of Mechanical Engineering Georgia Institute of Technology Atlanta GA USA Key Laboratory of Measurement and Control of Complex System of Engineering School of Automation Southeast University Nanjing China

ISBN: (数字)9798331518493

ISBN: (纸本)9798331518509

This paper introduces an adaptive robust control approach tailored for underactuated mechanical systems encountering matched and mismatched uncertainty, employing a constraint-following methodology. The control strategy unfolds in two phases: initially, a nominal control scheme is devised neglecting uncertainty and deviations in initial conditions from constraints. Subsequently, uncertainty is categorized into matched and mismatched components, ensuring that mismatched uncertainties remain unobservable. Leveraging the structural characteristics of the uncertainty bound, a novel segmented adaptive law is proposed and seamlessly integrated into the adaptive robust control framework. By employing the Lyapunov minimax approach, the method ensures uniform boundedness and uniform ultimate boundedness simultaneously, thereby ensuring approximate adherence to constraints for underactuated mechanical systems facing both matched and mismatched uncertainties alongside initial condition deviations.

关键词： Robust control Uncertainty Adaptive systems Costs Automation Nonlinear dynamical systems Mechanical systems Closed loop systems Robots

来源：评论

学校读者我要写书评

暂无评论

A High-Quality View Synthesis System for Sparse Rgb-D Input

SSRN

引用

SSRN 2024年

作者： Zhao, Ziyu Lu, Xiaobo Zhou, Sifan Chiang, Patrick Y.I.N. School of Automation Southeast University Nanjing China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China Key Laboratory of ASIC and System Fudan University Shanghai China

Recent research in neural rendering has achieved impressive results in the field of novel view synthesis. However, high-quality view synthesis with sparse source views remains a challenging task, as the reduction in the number of source views significantly degrades the synthesis quality. In this paper, we present a novel view synthesis system that exploits RGB-D pairs from sparse source views. First, we introduce a two-branch feature extraction network to embed the geometric clues from depth maps into the texture information of color images and fuse it with regularized Fourier mapping features. Second, a distance-aware feature query method, which constructs multi-level feature maps, is proposed to ensure high-fidelity details across different distance views. Then, we apply an occupancy network to perform occupancy prediction on raw point cloud. By integrating occupancy probabilities and extracted features, we use a prediction network with geometric information as an additional input to obtain the color and density of sampled points, and synthesize the target view using neural volume rendering. To prevent model degradation and enhance efficiency, we design an efficient sampling strategy based on Gaussian Mixture Model. Moreover, we introduce pixel-wise local regularization and full-image global geometric regularization to optimize the model. Extensive experiments show that the proposed method can achieve the state-of-the-art results for novel view synthesis with sparse inputs, and the model performance has been significantly improved in different scenarios. © 2024, The Authors. All rights reserved.

关键词： Geometry

来源：评论

学校读者我要写书评

暂无评论

MALT: Multi-scale Action Learning Transformer for Online Action Detection

MALT: Multi-scale Action Learning Transformer for Online Act...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Zhipeng Yang Ruoyu Wang Yang Tan Liping Xie The Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education School of Automation Southeast University Nanjing China Southeast University Shenzhen Research Institute Shenzhen China

ISBN: (数字)9798350359312

ISBN: (纸本)9798350359329

Online action detection (OAD) aims to identify ongoing actions from streaming video in real-time, without access to future frames. Since these actions manifest at varying scales of granularity, ranging from coarse to fine, projecting an entire set of action frames to a single latent encoding may result in a lack of local information, necessitating the acquisition of action features across multiple scales. In this paper, we propose a multi-scale action learning transformer (MALT), which includes a novel recurrent decoder (used for feature fusion) that includes fewer parameters and can be trained more efficiently. A hierarchical encoder with multiple encoding branches is further proposed to capture multi-scale action features. The output from the preceding branch is then incrementally input to the subsequent branch as part of a cross-attention calculation. In this way, output features transition from coarse to fine as the branches deepen. We also introduce an explicit frame scoring mechanism employing sparse attention, which filters irrelevant frames more efficiently, without requiring an additional network. The proposed method achieved state-of-the-art performance on two benchmark datasets (THUMOS’14 and TVSeries), outperforming all existing models used for comparison, with an mAP of 0.2% for THUMOS’14 and an mcAP of 0.1% for TVseries.

关键词： Filters Neural networks Streaming media Transformers Feature extraction Encoding Real-time systems

来源：评论

学校读者我要写书评

暂无评论

Stabilization of Port-controlled Hamiltonian systems subject to exogenous disturbances via compensation control approach

Stabilization of Port-Controlled Hamiltonian systems subject...

引用

Data Driven control and Learning systems (DDCLS)

作者： Baozeng Fu Qingzhi Wang Yongchao Liu Shandong Key Laboratory of Industrial Control Technology School of Automation Qingdao University Qingdao China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China

The stabilization of Port-controlled Hamiltonian (PCH) systems under exogenous disturbances by designing a compensation control approach is studied in the paper. Firstly, a baseline feedback control is proposed utilizing the damping injection method. To estimate disturbances, then a novel exogenous disturbance observer is constructed, based on which a feedforward compensation control is developed. Next, a composite control law is proposed based on the feedforward compensation and the feedback control. For the Hamiltonian system, an asymptotic stability analysis is followed finally. The feasibility of the compensation control strategy is revealed by a numerical simulation example.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A long prediction horizon fcs-mpc algorithm with disturbance preview for three-phase inverters

A long prediction horizon fcs-mpc algorithm with disturbance...

引用

2023 Jiangsu Annual Conference on Automation, JACA 2023

作者： He, Xianghui Huang, Saijin Li, Guanjun Li, Haoyuan Wang, Xiangyu School of Automation Southeast University No. 2 Si Pai Lou Nanjing Jiangsu210096 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education China Jiangsu Engineering Technology Research Center for Energy Storage Conversion Application China Electric Power Research Institute Nanjing Jiangsu210003 China

ISBN: (纸本)9781839539848

This paper derives an efficient control algorithm to deal with the output voltage tracking control problem of threephase inverters with multiple disturbances (including parametric perturbations of filter, abrupt disturbances caused by load switching and harmonic disturbances brought by unbalanced loads and nonlinear loads). This accomplishment is realized through the long prediction horizon finite-control-set model predictive control (FCS-MPC) and the integration of a composite disturbance observer (CDO) and a disturbance preview model into the sphere decoding algorithm (SDA). This integration facilitates the estimation and prediction of future disturbance behaviors within the inverter system, which addresses the poor disturbance-rejection limitations inherent in conventional SDA approaches employed in previous studies. The algorithm also improves the dynamic performances of the output voltage. Simulations studies are illustrated to demonstrate the feasibility and effectiveness of the proposed control algorithm. Based on the presented results, the algorithm achieves good disturbance rejection capability and maintains small tracking error. © The Institution of engineering & Technology 2023.

关键词： Disturbance rejection

来源：评论

学校读者我要写书评

暂无评论

MADDPG: Multi-agent Deep Deterministic Policy Gradient Algorithm for Formation Elliptical Encirclement and Collision Avoidance 5th

MADDPG: Multi-agent Deep Deterministic Policy Gradient Algor...

引用

5th Chinese Conference on Swarm Intelligence and Cooperative control, CCSICC 2021

作者： Xu, Leixin Chen, Weibin Liu, Xiang Chen, Yang-Yang School of Automation Southeast University Nanjing210096 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing210096 China Jiangsu Automation Research Institute Lianyungang222061 China China Tiesiju Civil Engineering Group Co. Ltd. Hefei230023 China

ISBN: (纸本)9789811939976

This paper considers the formation encirclement problem of multi-agent systems with collision avoidance, where the enclosing orbit is an ellipse, the target and the formation are dynamic. The novel rewards, that is the elliptical encirclement reward, the formation reward, the angular velocity reward and collision avoidance reward are designed and a reinforcement learning (RL) algorithm, that is multi-agent deep deterministic policy gradient (MADDPG), is designed based on the novel setting of rewards. Simulation results are given to show the validity of the proposed method. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

A Nonsmooth Observer Based on Friction Model Compensation for Sensorless Permanent Magnet Synchronous Motor

A Nonsmooth Observer Based on Friction Model Compensation fo...

引用

第42届中国控制会议

作者： Zaishan Zhang Wenwu Zhu Ke Yu Xiangyu Wang Shihua Li School of Automation Southeast University Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education School of Electrical Engineering and Automation Hefei University of Technology

In this paper,a nonsmooth observer considering friction compensation is proposed for the sensorless permanent magnet synchronous motor(PMSM) to estimate the rotor position and *** is one of the main factors affecting the tracking accuracy of the motor at low speed,especially in the back and forth motion of the ***,based on the Stribeck friction model,a nonlinear friction model of the motor is ***,a nonsmooth observer with friction compensation is designed to estimate the position and speed of the rotor,and the lumped disturbance of the system at the same ***,the stability analysis of this nonsmooth observer is ***,the effectiveness of the proposed method is verified by the simulation results.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Single-Domain Generalization Combining Geometric Context Toward Instance Segmentation of Track Components

Single-Domain Generalization Combining Geometric Context Tow...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Shixiang Su Songlin Du Dezhou Wang Shuzhen Tong Xiaobo Lu School of Automation Southeast University Nanjing China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing China School of Instrument Science and Engineering Southeast University Nanjing China

ISBN: (数字)9798350359312

ISBN: (纸本)9798350359329

A stable track components segmentation model should have consistent performance across a broad spectrum of railroad conditions, particularly in unfamiliar locations. Despite this, satisfying this requirement proves challenging when working with a limited track dataset, as there is a substantial domain shift between the given dataset and unobserved distributions. The goal of this paper is to improve the generalization ability of track component segmentation in situations where single-domain training data is available. Toward this end, a novel track component instance segmentation method combining the geometric context is proposed. First, we design an initial mask prediction head (IMPH) that utilizes predicted box output from the object detector to generate initial masks by merging the geometric priors of track components. Meanwhile, a multiscale feature fusion structure is introduced to encourage IMPH to better capture the geometric context. Then, a final mask refinement head (FMRH) is introduced to get higher quality masks with a geometry-gated aggregation strategy. On the basis of experiments on track datasets, it was determined that the heads can be incorporated with several types of detection frameworks and have demonstrated consistent generalization enhancements across multiple object detectors. Furthermore, our method substantially enhances segmentation performance on various unobserved domains.

关键词： Instance segmentation Geometry Accuracy Neural networks Merging Training data Detectors

来源：评论

学校读者我要写书评

暂无评论

Smoky Vehicle Detection Based on Improved Vision Transformer 21

Smoky Vehicle Detection Based on Improved Vision Transformer

引用

5th International Conference on Computer Science and Application engineering, CSAE 2021

作者： Yuan, Li Tong, Shuzhen Lu, Xiaobo School of Automation Southeast University Nanjing China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing China

ISBN: (纸本)9781450389853

The harmful exhaust emissions of fuel vehicles in theworld are damaging to human health and the environment, thus detecting smoky vehicles from real road environment is significant. At present, methods of smoky vehicle detection based on deep learning have the problem of high false-positive rate. To improve the performance, a two-stage video smoky vehicle detection algorithm based on the smoke classification in the core region from detected vehicle object boxes is proposed in this paper. Specifically, the vehicle object detection is realized by the algorithm based on YOLOv3. The smoke classification is realized by combining Vision Transformer and distillation, and the loss function is optimized in the training process. Experimental results on our smoky vehicle dataset have shown that the improved model achieves an F1 score over 0.4, precision over 0.4, recall nearly 0.1 improvement compared with the basic model, which can effectively reduce the false-positive rate during detection.. © 2021 Association for Computing Machinery. All rights reserved.

关键词： Smoke

来源：评论

学校读者我要写书评

暂无评论

A Delay-free control Method Based On Function Approximation And Broadcast For Robotic Surface And Multiactuator systems

arXiv

引用

arXiv 2024年

作者： Zhao, Yuchen School of Automation Southeast University Jiangsu Nanjing210096 China Ministry of Education Key Laboratory of Measurement and Control of Complex Systems of Engineering Southeast University Jiangsu Nanjing210096 China

Robotic surface consisting of many actuators can change shape to perform tasks, such as facilitating human-machine interactions and transporting objects. Increasing the number of actuators can enhance the robot's capacity, but controlling them requires communication bandwidth to increase equally in order to avoid time delays. We propose a novel control method that has constant time delays no matter how many actuators are in the robot. Having a distributed nature, the method first approximates target shapes, then broadcasts the approximation coefficients to the actuators, and relies on themselves to compute the inputs. We build a robotic pin array and measure the time delay as a function of the number of actuators to confirm the system size-independent scaling behavior. The shape-changing ability is achieved based on function approximation algorithms, i.e. discrete cosine transform or matching pursuit. We perform experiments to approximate target shapes and make quantitative comparison with those obtained from standard sequential control method. A good agreement between the experiments and theoretical predictions is achieved, and our method is more efficient in the sense that it requires less control messages to generate shapes with the same accuracy. Our method is also capable of dynamic tasks such as object manipulation. © 2024, CC BY.

关键词： Time delay

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：