检索结果-内蒙古大学图书馆

20th IEEE International Conference on Mechatronics and Automation, ICMA 2023

作者： Li, Chenqi Guo, Jian Guo, Shuxiang Fu, Qiang Tianjin University of Technology Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems Intelligent Robot Laboratory BinshuiXidao Extension 391 Tianjin300384 China Shenzhen Institute of Advanced Biomedical Robot Co. Ltd. No.12 Ganli Sixth Road Jihua Street Longgang District Shenzhen518100 China Key Laboratory of Convergence Medical Engineering Systems and Healthcare Technology The Ministry of Industry and Information Technology School of Life Science Beijing Institute of Technology No.5 Zhongguancun South Street Beijing100081 China

ISBN: (纸本)9798350320831

A path planning method based on a two-layer coordination system is proposed for the compatibility requirement of path planning global optimum and local coordination in a spherical multi-robot system. In the first layer of the algorithm, the global path quality is improved by proposing a nonlinear sine factor and a sparrow search algorithm with improved levy flight strategy to plan the global collision-free optimal path for each robot. In the second layer, a rolling window method is used to determine the possible collision positions of each robot in the system, and a priority-based conflict resolution strategy is proposed to complete the local coordination of the path near the corresponding position. The experimental results show that the path planning method enables the robots to drive along the global optimal path, but still can carry out flexible and orderly local path coordination, which effectively improves the path planning performance of the system. © 2023 IEEE.

关键词： Multipurpose robots

来源：评论

学校读者我要写书评

暂无评论

Multi-Feature Fusion Transformer for Natural Language Inference

Multi-Feature Fusion Transformer for Natural Language Infere...

引用

Robotics, Intelligent control and Artificial Intelligence (RICAI), International Conference on

作者： Lei Sun Xiangyang Li Hengxin Yan Tianj in Key Laboratory for Control Theory and Complicated Industry Systems School of Electrical Engineering and Automation Tianjin University of Technology Tianjin China School of Electrical Engineering and Automation Tianjin University of Technology Tianjin China

Natural Language Inference (NLI) is a branch of Natural Language Processing (NLP) whose main task is to determine the relationship between two sentences. Such tasks essentially use pre-trained models to ensure accuracy, and many applications are found in human-computer dialog and question-answering systems. However, in some cases where small devices are deployed offline, lightweight model implementations are often required to save computational resources. To address this problem and ensure more efficient inference capabilities, this paper describes how to improve the preexisting Transformer and proposes a Multi-Feature Fusion Transformer Network for NLI processing. The model integrates sequence features and introduces a priori information to enhance local features. It also achieves a comprehensive fusion of sequence information, local features, and non-local features, thus compensating for many potential shortcomings of Transformer. At the same time, it maintains the lightweight feature of the model and facilitates the implementation of downstream tasks. We verified the above point by experiments, our model has better inference performance compared to previous models used for the same task. The accuracy in the SNLI dataset reached 89.0%, while the matched and mismatched versions of MultiNLI reached 79.3% and 78.7%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Path Planning of Mobile Robot Based on Improved Artificial Potential Field Method

Path Planning of Mobile Robot Based on Improved Artificial P...

引用

Chinese control and Decision Conference, CCDC

作者： Jianyun Ni Helei Du Tie Wang Hao Li Chenyang Xue Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems Tianjin China School of Electronic Engineering and Automation Tianjin University of Technology Tianjin China Tianjin Climate Center Tianjin China Tianjin University of Technology Tianjin China

An improved artificial potential field method is proposed to address the problems of target unreachability and falling into local minima in the path planning process of the traditional artificial potential field method. Firstly, the distance factor between the robot and the target point is introduced to solve the target unreachability problem; secondly, the robot is guided out of the local minima point by setting the virtual target point through the double-circle strategy; finally, in order to satisfy the continuity of robot velocity and acceleration, the resulting path is smoothed by using three uniform B-sample curves. The experimental results show that the improved algorithm can effectively solve the problems of target unreachability and falling into local minima in the traditional algorithm, and the smoothness of the path after the spline treatment is better compared with the original path.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Grade quantification of Parkinson's disease based on principal component analysis

Grade quantification of Parkinson's disease based on princip...

引用

2021 China Automation Congress, CAC 2021

作者： Zhou, Long Tong, Jigang Dong, Enzeng Zhu, Lin Zhang, Jiachen Tianjin Key Laboratory For Control Theory and Applications in Complicated Systems Tianjin University of Technology Tianjin300384 China China North Industries Group 210 Reseasrch Institute Beijing China

ISBN: (纸本)9781665426473

Parkinson's disease is a common neurodegenerative disease, which seriously affects the life of patients. Whether there is an objective and reliable method to classify and identify the disease grade is of great significance in the clinical diagnosis of Parkinson's disease. In this paper, principal component analysis was used to comprehensively evaluate the degree of Parkinson's disease of 28 randomly selected subjects, establish a comprehensive scoring model of Parkinson's disease degree, and fit it according to HY grade, so as to realize the quantification of Parkinson's disease grade. Based on the data of vertical ground reaction force, 12 gait features with strong correlation are calculated as the main features to establish the scoring model. The validity of the model is verified by predicting the HY grade of another 12 subjects in the randomly selected data set. The results show that the prediction accuracy is 83.3%, which can effectively predict the Parkinson's disease level of subjects. © 2021 IEEE

关键词： Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

PID control for Aeroengine Based on Sparrow Search Algorithm

PID Control for Aeroengine Based on Sparrow Search Algorithm

引用

2021 China Automation Congress, CAC 2021

作者： Xu, Liang Wang, Hao Liu, Ying Xue, Wei Li, Ningning Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems School of Electrical Engineering and Automation Tianjin University of Technology Tianjin China National Key Laboratory for Complex Systems Simulation Beijing China BeijingAerospace Propulsion Institute Beijing China

ISBN: (纸本)9781665426473

To address the problem of aeroengine PID parameter optimization and tuning, this work employs the excellent performance of the sparrow search algorithm (SSA) with high convergence accuracy and fast convergence speed to propose a self-tuning PID control method (SSA-PID) for aeroengine parameters. Considering that the oil supply output by the actuator may have a substantial overshoot, the proposed technique adds a penalty mechanism for the overshoot at the output, designs and improves the fitness function weighted by performance indicators, and effectively avoids the possible excessive overshoot and sharp oil supply phenomenon of the engine oil supply. The SSA-PID simulates and verifies the mathematical models of the turbofan engine ' s steady-state operating points on the ground and at high altitudes. The results show that the SSA-PID control method has better parameter tuning and optimization effects, a faster speed step response, and generates a fuel supply that does not have a large overshoot, thereby proving its effectiveness and feasibility. © 2021 IEEE

关键词： Three term control systems

来源：评论

学校读者我要写书评

暂无评论

Dual-beam photothermal material trace detection method based on self-mixing interference

引用

Applied Optics 2025年第16期64卷 4549-4558页

作者： Yan Zhao Yunlong Zhi Hui Zhao Daming Dong Jigang Tong Chunling Tu School of Electrical Engineering and Automation Tianjin University of Technology Tianjin 300384 China Tianjin Key Laboratory for Control Theory and Applications in Complicated Systems Tianjin 300384 China National Engineering Research Center for Intelligent Agricultural Equipment Beijing Academy of Agriculture and Forestry Sciences Beijing 100097 China Department of Computer Systems Engineering Faculty of ICT Tshwane University of Technology Pretoria 0188 South Africa

In this paper, a dual-beam photothermal self-mixing substance trace detection system is proposed. The crystal violet (CV) solution of the sample undergoes a thermal lens effect under the action of pump photoperiod excitation, and the refractive index changes cause the deviation of the probe beam, and the self-mixing interference (SMI) effect occurs when the modulated light returns to the cavity. By combining theoretical and experimental results, the power fluctuation of the SMI signal caused by the thermo-optic effect is linearly related to CV concentration and other parameters. At an excitation power of 24 mW, the detection limit can reach $8.0 \times {10^{- 8}}\;{\rm mol/L}$ . This method does not require labeling or complex CV pretreatment and has high sensitivity and flexibility, providing a guide for CV characterization in biological, environmental, and pharmaceutical research.

关键词： Gaussian beams Laser beams Optical fields Phase shift Refractive index Spatial resolution

来源：评论

学校读者我要写书评

暂无评论

Motor Imagery and Mental Arithmetic Classification Based on Transformer Deep Learning Network

Motor Imagery and Mental Arithmetic Classification Based on ...

引用

IEEE International Conference on Mechatronics and Automation

作者： Yuanyuan Ye Jigang Tong Sen Yang Yinghui Chang Shengzhi Du Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory Tianjin University of Technology Tianjin China National Clinical Research Center for Chinese Medicine Acupuncture and Moxibustion Tianjin China Department of Electrical Engineering Tshwane University of Technology Pretoria South Africa

ISBN: (数字)9798350388077

ISBN: (纸本)9798350388084

Non-invasive electroencephalography (EEG) signals find widespread application in brain-computer interfaces (BCI), with paradigms such as motor imagery (MI), mental arithmetic (MA) and emotion recognition being particularly common. This study aims to explore feature extraction and classification methods for MI and MA tasks. We employed convolutional neural networks (CNN) combined with Transformer networks and Fasternet Block to extract features. Through our reaearch, we obtained features for MI and MA tasks, and used softmax classification for binary classification of these tasks. We conducted on a publicly available dataset consisting of data from 29 subjects. The experimental results demonstrate that our method achieved high classification accuracy in MI and MA tasks. The final accuracy rates for MI and MA tasks were 88.67% and 91.23%, respectively.

关键词： Accuracy Mechatronics Feature extraction Brain modeling Transformers Motors Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

EEG Signal Driving Fatigue Detection based on Differential Entropy

EEG Signal Driving Fatigue Detection based on Differential E...

引用

IEEE International Conference on Mechatronics and Automation

作者： Danyang Wang Jigang Tong Sen Yang Yinghui Chang Shengzhi Du Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory Tianjin University of Technology Tianjin China National Clinical Research Center for Chinese Medicine Acupuncture and Moxibustion Tianjin China Department of Electrical Engineering Tshwane University of Technology Pretoria South Africa

ISBN: (数字)9798350388077

ISBN: (纸本)9798350388084

With the gradual increase of vehicle penetration rate, driving safety has become a problem that people pay attention to, and traffic accidents caused by fatigue are mostly, therefore, it is meaningful to devote ourselves to research on the method of detecting driving fatigue so as to achieve safety warning. Based on this, in this paper, we collect the EEG signals of the experimenter through driving simulation, then extract the features through differential entropy, classify the features with GBDT, and finally determine the fatigue level of the subject with the classification results. We use our method to analyse the difference with sample entropy and fuzzy entropy at the feature level, and with KNN and SVM at the classification level. The comparison shows that our method has significant accuracy in EEG-based fatigue *** can be used accurately in extraction methods through its properties, and differential entropy-based GBDT methods may be useful in detecting driver fatigue.

关键词： Support vector machines Mechatronics Nearest neighbor methods Fatigue Feature extraction Entropy Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

Attention-Based Grayscale Image Colorization

Attention-Based Grayscale Image Colorization

引用

IEEE International Conference on Mechatronics and Automation

作者： Lei Sun Dongjie Ju Shi Ke School of Electrical Engineering and Automation Binshui Xidao Extension 391 Tianjin China Tianjin Key Laboratory for Control Theory and Complicated Industry Systems Tianjin University of Technology Binshui Xidao Extension 391 Tianjin China

Image colorization has always been a hot topic in computer vision. Since the emergence of deep learning and its excellent performance in many image-processing tasks, image colorization methods based on convolutional neural networks (CNN) emerge in an endless stream, including automatic colorization methods and A coloring method based on a reference image. In the coloring method based on the reference image, an existing problem is that an algorithm for calculating the similarity between the grayscale image and the reference image needs to be designed to map the color blocks in the reference image to the grayscale image. And we use the attention mechanism to spread color implicitly, and propose an attention-guided image colorization network, which extracts features and corresponding color information from the reference image through a reference encoding sub-network, and simultaneously passes a target encoding sub-network The network extracts grayscale features from grayscale images, uses attention to calculate the correlation between the two types of features, and then restores the features and reconstructs colors through the decoding network. Our method can quickly capture the color of the reference feature, accurately transfer it to the target feature, and effectively solve the problem of small target loss during downsampling. By validating on multiple datasets, the results show that our method is effective, outperforming multiple state-of-the-art image colorization algorithms.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multilevel Decomposition Time Aggregation Graph Convolution Networks for Skeleton-Based Action Recognition

Multilevel Decomposition Time Aggregation Graph Convolution ...

引用

IEEE International Conference on Mechatronics and Automation

作者： Wenhua Li Enzeng Dong Jigang Tong Sen Yang Zufeng Zhang Wenyu Li Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems and Intelligent Robot Laboratory Tianjin University of Technology 391 Binshui West Road Tianjin China College of Artificial Intelligence Nankai University 94 Weijin Road Tianjin China

Skeleton-based human action recognition has become a popular topic among researchers. This is because using skeletal data provides a robust solution to problems encountered in complex environments, such as changes in perspective and background interference. The robustness of skeletal data enables recognition methods to focus on more specific and relevant features. We propose a model called multilevel decomposition time aggregation graph convolution network (MDT-GCN), which utilizes a multilevel graph convolution kernel to capture higher-order spatial dependence relationships between joints. This is achieved by decomposing a human topology graph into smaller graphs, each of which has its own graph convolution kernel. To further enhance the performance of our model, we employ a two-flow framework and channel topology refinement strategy. Our experiments on the NTU-RGB+D60 and NTU-RGB+D120 datasets demonstrate that our MDT-GCN network outperforms the previous algorithm and significantly improves the accuracy of action recognition.

关键词：

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：