检索结果-内蒙古大学图书馆

YOLO-MS: Multispectral Object Detection via Feature Interaction and Self-Attention Guided Fusion

IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL systems 2023年第4期15卷 2132-2143页

作者： Xie, Yumin Zhang, Langwen Yu, Xiaoyuan Xie, Wei South China Univ Technol Coll Automat Sci & Technol Guangzhou 510640 Peoples R China South China Univ Technol Coll Automat Sci & Technol Key Lab Autonomous Syst & Networked Control Minist Educ Guangzhou 510641 Peoples R China South China Univ Technol Unmanned Aerial Vehicle Syst Engn Technol Res Ctr Guangzhou 510641 Peoples R China South China Normal Univ Coll Phys & Telecommun Engn Guangzhou 510006 Peoples R China

Object detection is essential for an autonomous driving sensing system. Since the light condition is changed in unconstrained scenarios, the detection accuracy based on visible images can be greatly degraded. Although the detection accuracy can be improved by fusing visible and infrared images, existing multispectral object detection (MOD) algorithms suffer from inadequate intermodal interaction and a lack of global dependence in the fusion approach. Thus, we propose an MOD framework called YOLO-MS by designing a feature interaction and self-attention fusion network (FISAFN) as the backbone network. Within the FISAFN, correlations between two modalities are extracted by the feature interaction module (FIM) for reconstructing the information components of each modality and enhancing capability of information exchange. To filter redundant features and enhance complementary features, long-range information dependence between two modalities are established by using a self-attention feature fusion module (SAFFM). Thus, a better information richness of the fused features can be achieved. Experimental results on the FLIR-aligned data set and the M3FD data set demonstrate that the proposed YOLO-MS performs favorably against state-of-the-art approaches, including feature-level fusion and pixel-level fusion. And further, the proposed YOLO-MS possesses good detection performance under diverse scene conditions.

关键词： Feature interaction multispectral features fusion object detection self-attention

来源：评论

学校读者我要写书评

暂无评论

Predefined-Time Attitude Tracking control for Quadrotor UAV with Input Delay

Predefined-Time Attitude Tracking Control for Quadrotor UAV ...

引用

Chinese Automation Congress (CAC)

作者： Shuo Li Hailong Pei Na Duan Key Laboratory of Autonomous Systems and Networked Control Ministry of Education Guangdong Engineering Technology Research Center of Unmanned Aerial Vehicle Systems College of Automation Science and Enginneering South China University of Technology Guangzhou China Key Laboratory of Autonomous Systems and Networked Control Ministry of Education Guangdong Engineering Technology Research Center of Unmanned Aerial Vehicle Systems School of Automation Science and Enginneering South China University of Technology Guangzhou China School of Electrical Engineering & Automation Jiangsu Normal University Xuzhou China

In this paper, the predefined time attitude tracking flight problem for a quadrotor unmanned aerial vehicle (UAV) system with input delays is investigated. By means of a novel auxiliary system, the effects of input delay are solved. Then, on the basis of the backstepping method and predefined performance control technology, the predefined-time attitude tracking controller is constructed to make UAV attitude tracking. Simulation results verify the validity of the proposed scheme.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multi-step and multi-task learning to predict quality-related variables in wastewater treatment processes

引用

PROCESS SAFETY AND ENVIRONMENTAL PROTECTION 2023年 180卷 404-416页

作者： Liu, Yiqi Yuan, Jingyi Cai, Baoping Chen, Hongtian Li, Yan Huang, Daoping South China Univ Technol Sch Automat Sci & Engn Minist Educ Key Lab Autonomous Syst & Networked Control Guangzhou 510640 Peoples R China South China Univ Technol Sch Automat Sci & Engn Guangdong Engn Technol Res Ctr Unmanned Aerial Veh Guangzhou 510640 Peoples R China China Univ Petr Coll Mech & Elect Engn Qingdao 266580 Shandong Peoples R China Shanghai Jiao Tong Univ Dept Automat Shanghai 200240 Peoples R China

In wastewater treatment processes, lack of hardware sensors together with unacceptable dynamics, strong nonlinearity and large time delay often leads to a large number of key variables that are difficult to measure online accurately and timely, then frustrate safe operations of the processes. To accurately and timely capture the short-term behavior changes and trend development of critical variables, a novel neural network based softsensing model is proposed to take full use of multi-task learning, direct multi-step prediction strategy and evolutionary algorithm to formulate a novel multi-task multi-step evolution (MTMSE) neural network. Firstly, single-output MTMSE (SO-MTMSE) neural network is used to realize the dynamic monitoring of a single variable. Moreover, by considering the spatiotemporal interaction among the data, the model is extended to multi-output MTMSE (MO-MTMSE) neural network to simultaneously realize multi-step prediction of multiple variables, thus providing a desired reference for optimizing the wastewater treatment processes. Finally, the proposed model is applied to the benchmark simulation model 2 (BSM2) and a full-scale wastewater treatment plant in Shenzhen. And the results show that the proposed dynamic soft sensor model outperforms the standard methods, such as autoregressive moving average model (ARMA) and multiple output gaussian process regression (MGPR).

关键词： Wastewater treatment Multi -step prediction Multi -task learning Soft sensor Neural network

来源：评论

学校读者我要写书评

暂无评论

A robotic learning and generalization framework for curved surface based on modified DMP

引用

ROBOTICS AND autonomous systems 2023年 160卷

作者： Xue, Xianfa Dong, Jiale Lu, Zhenyu Wang, Ning South China Univ Technol Sch Automation Sci & Engn Key Lab Autonomous Syst & Networked Control Guangzhou Peoples R China Univ West England Bristol Robot Lab Coldharbour Ln Frenchay BS16 1QY England

How to reproduce and generalize the skills acquired by demonstrating is a hot topic for researchers. (1) A compliant continuous drag demonstration system based on discrete admittance model was designed to continuously and smoothly drag or demonstrate. (2) The modified DMP including the scaling factor and the force coupling term was used to improve the poor generalization ability of the classical DMP. (3) Curve drawing experiments were carried out to show the effectiveness of our proposed learning and generalization framework.(c) 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://***/licenses/by/4.0/).

关键词： Continuous drag demonstration Modified DMP Curve drawing experiments Learning and generalization framework

来源：评论

学校读者我要写书评

暂无评论

Design and control of a Novel 5-DoF Tilt-Rotor Quadcopter

Design and Control of a Novel 5-DoF Tilt-Rotor Quadcopter

引用

Chinese Automation Congress (CAC)

作者： Zhenwei Lin Wei Liu Dabo Xu Key Laboratory of Autonomous Systems and Networked Control of Ministry of Education School of Automation Science and Engineering South China University of Technology Guangzhou China Shien-Ming Wu School of Intelligent Engineering South China University of Technology Guangzhou China

ISBN: (数字)9798350368604

ISBN: (纸本)9798350368611

This paper presents a novel 5-DoF (degrees of freedom) tilt-rotor quadcopter. Compared to traditional quadcopters, the vehicle offers more DoF by adding two servo-driven mechanisms. The extra DoF allow it to move at a constant pitch angle, thus adapting it to more scenarios, such as slope landings. In addition, the proposed tilt-rotor quadcopter can generate greater torque around the z-axis of the body coordinate system. This gives it an advantage in some scenarios, such as carrying a manipulator to turn valves in the air. In this paper, we first design and model the proposed tilt-rotor quadcopter. Then, we design the controller and control allocation algorithm for the proposed tilt-rotor quadcopter. The control allocation algorithm has clear analytical solutions, which benefits its implementation and deployment. Finally, the structure and control strategy of the tilt-rotor quadcopter are validated by an experiment.

关键词： Torque Automation 5-DOF Atmospheric modeling Valves Manipulators Resource management Quadrotors

来源：评论

学校读者我要写书评

暂无评论

Study on High-Speed-to-Hovering Back-Transition control of Ducted Fan UAV 33

Study on High-Speed-to-Hovering Back-Transition Control of D...

引用

33rd Chinese control and Decision Conference (CCDC)

作者： Chen, Feng Pei, Hailong Cheng, Zihuan South China Univ Technol Unmanned Aerial Vehicle Syst Engn Technol Res Ctr Minist Educ Key Lab Autonomous Syst & Networked Control Guangzhou 510640 Peoples R China

ISBN: (纸本)9781665440899

The ducted fan unmanned aerial vehicle (UAV) can transit from hovering to high-speed flight by continuously pitching down, reaching a high-speed and high-efficiency status. However, during the back-transition process from high speed to hovering, unstable environmental airflow and improper velocity planning could lead to serious oscillation, or even divergence of the aircraft states. In this paper, we address this control problem of a ducted fan UAV by utilizing an expected velocity planning method and a robust controller. The expected velocity curve is planned to ensure the dynamic effect of the deceleration process and decrease the computational complexity in solving the numerical optimization problem. The robust controller aims to reduce the tracking error of velocity and angular velocity considering modeling uncertainty and disturbance. Furthermore, the unit quaternion is selected to represent the attitude to avoid the singularity problem. Theoretical analysis verifies the stability of the proposed controller based on unit quaternion. Moreover, the simulation and experimental results collectively indicate the effectiveness of our methods.

关键词： Ducted Fan UAV Back-Transition Velocity Planning Robust control Unit Quaternion

来源：评论

学校读者我要写书评

暂无评论

Uncertainty and Reliability Analysis of Neural Networks Based Soft-sensors With Applications in Wastewater Treatment*

Uncertainty and Reliability Analysis of Neural Networks Base...

引用

Chinese Automation Congress (CAC)

作者： Yiqi Liu Jingyi Yuan Daoping Huang Key Laboratory of Autonomous Systems and Networked Control Ministry of Education the School of Automation Science & Engineering South China University of Technology Guangzhou China

To overcome the transparency and interpretability issues of neural network-based soft sensing, a data-driven framework by combining neural network prediction models and sensitivity analysis methods is constructed in this paper, with application in wastewater treatment plant analysis. The results show that readily biodegradable substrate effluent (SS-e) and ${\text{N}}{{\text{H}}_4}^ + + {\text{N}}{{\text{H}}_3}$ nitrogen effluent (SNH-e) are both important input factors affecting the difficult-to-measured effluent quality, chemical oxygen demand (COD), by resorting to PAWN sensitivity analysis and four neural network models. In addition, four neural network models worked together with bilateral sensitivity analysis for the input perturbation analysis, which shows that the addition of Gaussian perturbation to SS-e and flow rate influent (Q-in) has a greater effect on the prediction which results large fluctuations of output results. Finally, the effectiveness of this framework is confirmed based on above results.

关键词： Analytical models Uncertainty Sensitivity analysis Biological system modeling Perturbation methods Neural networks Effluents

来源：评论

学校读者我要写书评

暂无评论

Sampled-Data Semi-Global Practical Tracking for Uncertain Euler-Lagrange systems

Sampled-Data Semi-Global Practical Tracking for Uncertain Eu...

引用

Chinese Automation Congress (CAC)

作者： Kaiji Zeng Wei Liu Key Laboratory of Autonomous Systems and Networked Control Ministry of Education School of Automation Science and Engineering South China University of Technology Guangzhou China

We investigate a sampled-data tracking control for uncertain Euler-Lagrange systems in this paper. What makes our result interesting is that our system and exosystem encompass some uncertain parameters that belong to some compact domain that can be specified arbitrarily large. To handle the uncertainties in both the Euler-Lagrange systems and the exosystem, we need to resort to not only the adaptive control technique but also the internal model method. By proposing a sampled-data control law, we solve the problem with guaranteed tracking performance, that is, the tracking errors can converge exponentially, by any fast specified rate, to any small specified neighborhood of the origin.

关键词： Adaptation models Upper bound Uncertainty Automation control systems Adaptive control

来源：评论

学校读者我要写书评

暂无评论

Adaptive Stiffness control of Series Elastic Actuator Manipulators Based on Dynamic System 20

Adaptive Stiffness Control of Series Elastic Actuator Manipu...

引用

IEEE 20th International Conference on Automation Science and Engineering (CASE)

作者： Guo, Kunlin Zeng, Chao Si, Weiyong Wang, Ning Yang, Chenguang South China Univ Technol Sch Automat Sci & Engn Key Lab Autonomous Syst & Networked Control Guangzhou 510640 Peoples R China Univ Liverpool Dept Comp Sci Liverpool L69 3BX Merseyside England Univ Essex Sch Comp Sci & Elect Engn Wivenhoe Pk Colchester CO4 3SQ Essex England Univ West England Bristol Robot Lab Bristol BS16 1QY Avon England

ISBN: (纸本)9798350358513;9798350358520

In traditional control methods, Series Elastic Actuator (SEA) joint manipulators are limited by their hardware and can only perform simple tasks with low stiffness. To address this issue, we propose a stiffness adjustment control strategy for SEA manipulators based on Dynamic systems (DS) with good generalization performance. By enhancing the generalization capability of DS in complex tasks and combining posture control, the SEA manipulators can autonomously adjust control gains and postures according to task requirements to achieve higher stiffness at the end-effector. Experimental results involving the activation of air switches with different stiffness levels and installation angles have demonstrated the effectiveness of our proposed method. The results indicate that compared to traditional methods of adjusting the control gain and retraining DS for similar tasks, our approach exhibits superior generalization performance while maintaining endeffector stiffness during interactions.

关键词： End effectors

来源：评论

学校读者我要写书评

暂无评论

Robust Offline Imitation Learning Through State-level Trajectory Stitching

arXiv

引用

arXiv 2025年

作者： Wang, Shuze Mei, Yunpeng Cao, Hongjie Yuan, Yetian Wang, Gang Sun, Jian Chen, Jie National Key Lab of Autonomous Intelligent Unmanned Systems Beijing Institute of Technology Beijing100081 China Department of Control Science and Engineering Harbin Institute of Technology Harbin150001 China

Imitation learning (IL) has proven effective for enabling robots to acquire visuomotor skills through expert demonstrations. However, traditional IL methods are limited by their reliance on high-quality, often scarce, expert data, and suffer from covariate shift. To address these challenges, recent advances in offline IL have incorporated suboptimal, unlabeled datasets into the training. In this paper, we propose a novel approach to enhance policy learning from mixed-quality offline datasets by leveraging task-relevant trajectory fragments and rich environmental dynamics. Specifically, we introduce a state-based search framework that stitches state-action pairs from imperfect demonstrations, generating more diverse and informative training trajectories. Experimental results on standard IL benchmarks and real-world robotic tasks showcase that our proposed method significantly improves both generalization and performance. Videos are available on https://***/sbr. Copyright © 2025, The Authors. All rights reserved.

关键词： Trajectories

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：