检索结果-内蒙古大学图书馆

International Conference on Applied Superconductivity and Electromagnetic Devices, ASEMD

作者： Guodong Yu Shuqi Sun Junlong Li Lijun Xiao Yongxiang Xu Jibin Zou State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin China

Although IPM motor with 1/2 slot per phase per pole has a relatively less rotor loss in comparison with other fractional slot concentrated winding (FSCW), the problems of THD of back-EMF and cogging torque are worse, which limits its application in electric vehicles and other industry fields. This paper proposes an effective rotor-shape optimization method to suppress both THD of back-EMF and cogging torque for improving torque ripple problem. Different from the prevalent rotor optimization methods, the proposed rotor-shape optimization method associates both shape function optimization of rotor surface with PM shaping, which can effectively decrease optimization parameters, and enhance optimization efficiency. To clarify the optimization strategy, the back-EMF and cogging torque for this PM motor with 1/2 slot per phase per pole are analyzed firstly. Then, the proposed optimization method is presented in detail. Finally, the optimized IPM motor is prototyped and its experimental results verifies the effectiveness of the proposed optimization method.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Research and Introduction of the Process Flow of 3D-IC Thermo-Compression Intelligent Collective Bonding Equipment

Research and Introduction of the Process Flow of 3D-IC Therm...

引用

International Conference on (ICEPT) Electronic Packaging technology

作者： Bin Zhou Jianxin Qiao Longqiu Li Yunkang Su Dapeng Liu Haichuan Shi Wentao Huang State Key Laboratory of Robotics and System Key Laboratory for Microsystems and Microstructure Manufacturing Harbin Institute of Technology Harbin China School of Mechanical and Electrical Engineering Harbin Institute of Technology Harbin China

ISBN: (数字)9798350353808

ISBN: (纸本)9798350353815

Three-dimensional integrated (3D IC) technologies have attracted substantial academic interest due to their enhanced performance and reduced interfacial dimensions. Conventional 3D packaging for chips of four layers must be stacked and bonded four and three times, respectively, bringing about the low reliability of bumps and productivity. Additionally, the decreasing size and spacing of solder bumps present challenges for defect detection. This paper proposes a collective intelligent bonding equipment for flip-chip 3D packaging. The equipment includes a bonding head system, downward vision, chip shift table, upward vision, bonding zone, in-situ monitoring system, and substrate movement platform. The bonding head system, downward vision, chip shift table, upward vision, and bonding zone complete the collective bonding. The defect detection is accomplished by the in-situ monitoring system. The chips need to be stacked four times, and the time of bonding is once. Meanwhile, during the bonding process, the defects can be monitored in real-time by the in situ monitoring system based on active infrared technology. Four layers in chips can be simultaneously bonded, and alignment accuracy is ±15μm@3sigma. The breakage, misalignment, and other defects can be monitored during the bonding progress. The introduction of this equipment will enhance productivity while ensuring bump reliability. Moreover, its successful development holds significant importance for the advancement and manufacturing of semiconductor packaging equipment domestically.

关键词： Productivity Three-dimensional displays Machine vision Packaging Reliability Bonding Soldering

来源：评论

学校读者我要写书评

暂无评论

Optimizing Dataset Configuration for Improving Convolutional Neural Networks Performance in Chest X-ray Image Classification

Optimizing Dataset Configuration for Improving Convolutional...

引用

6th International Conference on Intelligent Computing and Signal Processing (ICSP)

作者： Zengmin He Yi Ren Honggang Wang State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin China

Chest X-ray (CXR) is essential for physicians to diagnose lung diseases in clinical medicine. With the development of computer and deep learning techniques, creating CXR datasets to train convolutional neural networks (CNNs) has become a popular research topic. However, there are still challenges in creating datasets due to class imbalance and the ratio setting of datasets. This study investigated the impact of class imbalance and the ratio of training, validation, and test sets on CNNs classification performance by optimizing the dataset configuration. This was achieved by directly modifying the ratio, applying oversampling based on the adaptive contrast enhancement (ACE) algorithm, and random undersampling to balance the dataset classes, followed by modifying the ratio again. Therefore, seven datasets were obtained, which were utilized to individually train four CNNs based on transfer learning and fine-tuning techniques. Evaluation metrics based on the confusion matrix were utilized to demonstrate the enhanced classification performance of CNNs. The results indicated that at least 17% of the accuracy of CNNs trained by the dataset with modified ratios was improved as compared with the dataset with an unreasonable initial configuration. Additionally, the overall evaluation metrics of CNNs were further improved by balancing the dataset classes and modifying the ratios. ChexNet demonstrated the best classification performance among the four CNNs, as evidenced by the area under the receiver operating characteristic (ROC) curves (AUC). Furthermore, ChexNet was trained by a class balanced dataset with a ratio of 7:2:1, resulting in the best evaluation metrics, including F1 Score, balanced accuracy, accuracy, recall, specificity, and precision, which were 97.81%, 97.78%, 97.78%, 98.73%, 96.84%, and 96.89%, respectively. Therefore, optimizing dataset configuration can effectively improve the performance of CNNs, providing empirical guidance for researchers in creating dat

关键词：

来源：评论

学校读者我要写书评

暂无评论

GCA-Net: A Global Context Aggregation Network for Effective Optical Flow

GCA-Net: A Global Context Aggregation Network for Effective ...

引用

Annual Conference of Industrial Electronics Society

作者： Tao Xie Jinghan Gao Ke Wang Ruifeng Li State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin China

Optical flow seeks to estimate the per-pixel 2D motion between two frames by identifying corresponding pixels. Current flow estimators typically involve per-pixel feature extraction, multi-scale 4D correlation volume construction, and iterative flow field updates through a Conv-GRU module. However, the locality of convolutional features in these methods renders the calculated correlations vulnerable to different noises. In addition, the Conv-GRU module of these methods is only executed with a convolution layer, which is incapable of exploiting context clues from larger window sizes even inside the image being queried itself, thus raising it more difficult for the model to process images with challenging regions, e.g., textureless areas. To this end, we introduce GCA-Net, a global context aggregation network for credible yet effective optical flow estimation. More specifically, we propose a highly efficient multi-scale transformer (MSFormer) layer which enables the per-pixel feature to aggregate long-range information from other features, hence building more accurate 4D correlation volumes. Besides, we develop an attention-enhanced Conv-GRU block (AttGRU) that can incorporate information alongside a larger context window even within itself empowering our network to estimate optical flow in even the most challenging regions. Experimentally, we demonstrate that GCA-Net outperforms previous state-of-the-art methods by large margins on Sintel (Final) and KITTI 2015 (background and foreground) benchmarks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Poly-MOT: A Polyhedral Framework For 3D Multi-Object Tracking

Poly-MOT: A Polyhedral Framework For 3D Multi-Object Trackin...

引用

IEEE/RSJ International Conference on Intelligent Robots and systems (IROS)

作者： Xiaoyu Li Tao Xie Dedong Liu Jinghan Gao Kun Dai Zhiqiang Jiang Lijun Zhao Ke Wang State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin China

3D Multi-object tracking (MOT) empowers mobile robots to accomplish well-informed motion planning and navigation tasks by providing motion trajectories of surrounding objects. However, existing 3D MOT methods typically employ a single similarity metric and physical model to perform data association and state estimation for all objects. With large-scale modern datasets and real scenes, there are a variety of object categories that commonly exhibit distinctive geometric properties and motion patterns. In this way, such distinctions would enable various object categories to behave differently under the same standard, resulting in erroneous matches between trajectories and detections, and jeopardizing the reliability of downstream tasks (navigation, etc.). Towards this end, we propose Poly-MOT, an efficient 3D MOT method based on the Tracking-By-Detection framework that enables the tracker to choose the most appropriate tracking criteria for each object category. Specifically, Poly-MOT leverages different motion models for various object categories to characterize distinct types of motion accurately. We also introduce the constraint of the rigid structure of objects into a specific motion model to accurately describe the highly nonlinear motion of the object. Additionally, we introduce a two-stage data association strategy to ensure that objects can find the optimal similarity metric from three custom metrics for their categories and reduce missing matches. On the NuScenes dataset, our proposed method achieves state-of-the-art performance with 75.4% AMOTA. The code is available at https://***/lixiaoyu20001P0Iy-MOT.

关键词：

来源：评论

学校读者我要写书评

暂无评论

HabitAction: A Video Dataset for Human Habitual Behavior Recognition

arXiv

引用

arXiv 2024年

作者： Li, Hongwu Zhang, Zhenliang Wang, Wei The State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin150080 China Beijing100080 China

Human Action Recognition (HAR) is a very crucial task in computer vision. It helps to carry out a series of downstream tasks, like understanding human behaviors. Due to the complexity of human behaviors, many highly valuable behaviors are not yet encompassed within the available datasets for HAR, e.g., human habitual behaviors (HHBs). HHBs hold significant importance for analyzing a person’s personality, habits, and psychological changes. To solve these problems, in this work, we build a novel video dataset to demonstrate various HHBs. These behaviors in the proposed dataset are able to reflect internal mental states and specific emotions of the characters, e.g., crossing arms suggests to shield oneself from perceived threats. The dataset contains 30 categories of habitual behaviors including more than 300,000 frames and 6,899 action instances. Since these behaviors usually appear at small local parts of human action videos, it is difficult for existing action recognition methods to handle these local features. Therefore, we also propose a two-stream model using both human skeletons and RGB appearances. Experimental results demonstrate that our proposed method has much better performance in action recognition than the existing methods on the proposed dataset. All the dataset and codes are available at ***/view/hhb30. Copyright © 2024, The Authors. All rights reserved.

关键词： Video analysis

来源：评论

学校读者我要写书评

暂无评论

A Flexible Piezoelectric-based Tactile Sensor for Dynamic Force Measurement 4

A Flexible Piezoelectric-based Tactile Sensor for Dynamic Fo...

引用

4th International Conference on High Performance Big Data and Intelligent systems, HDIS 2022

作者： Xiang, Guiyao Wang, Xiaoxin Cheng, Nengyuan Hu, Liang Zhang, Hongwei Liu, Honghai School of Mechanical Engineering and Automation Harbin Institute of Technology Shenzhen Shenzhen China Harbin Institute of Technology Shenzhen State Key Laboratory of Robotics and System Shenzhen China

ISBN: (数字)9781665491440

ISBN: (纸本)9781665491440

Force measurement is important in the robotic manipulation tasks, and it is an essential property of specific tactile sensor. An increasing number of studies demonstrated the capacity of tactile sensor in force measurement. However, most of the works focus on the static force measurement, the precise track of dynamic force is hard to be accomplished. Compared with traditional piezoresistive sensors, piezoelectric-based tactile sensors have advantages in low energy consumption and high sensitivity. This paper developed a flexible multi-layer piezoelectric-based tactile sensor inspired by the structure of skin. The size of sensor with a 3×3 sensing element array was designed to match that of a finger. The signal processors and logical algorithms were integrated to sense and distinguish the size, location and pattern of different external force stimuli in real time, including pressure intensity and vibration frequency. Two experiments about pressure sensing and vibration sensing show that the proposed tactile sensor with finger-like force range, high sensitivity and fast response time. The range of measured force amplitude and frequency were 0.1N-2N and 0.1Hz-100Hz. Another characteristic, the response time, was less than 50ms. In addition, the sensor array has a spatial resolution of 4 mm. This work has driven the development of tactile sensor for dynamic force measurement in real-Time robotic grasping. © 2022 IEEE.

关键词： Energy utilization

来源：评论

学校读者我要写书评

暂无评论

Research on Longitudinal Ultrasonic Vibration Assisted Flip Chip Bonding Process

Research on Longitudinal Ultrasonic Vibration Assisted Flip ...

引用

International Conference on (ICEPT) Electronic Packaging technology

作者： Dapeng Liu Jianxin Qiao Longqiu Li Zhi Qiao Bin Zhou Yunkang Su State Key Laboratory of Robotics and System Key Laboratory for Microsystems and Microstructure Manufacturing Harbin Institute of Technology Harbin China School of Mechanical and Electrical Engineering Harbin Institute of Technology Harbin China

ISBN: (数字)9798350353808

ISBN: (纸本)9798350353815

The current flip-chip bonding technology still faces defects such as surface thermal burnout of chips due to excessively high process temperature and significant mechanical stress leading to solder joint fracture and chip damage. This paper proposes a longitudinal ultrasonic-assisted flip-chip bonding process. By introducing ultrasonic vibration of a certain frequency in the bonding direction, stress on the device during the bonding process is reduced, leading to lower bonding temperatures and improved bonding quality and reliability. However, unreasonable ultrasonic bonding parameters may result in the fracture at the bonding interface by hammering effect. Based on the longitudinal ultrasonic-assisted bonding experimental platform was constructed to conduct bonding experiments. Through orthogonal contrast experiments, it has been determined that longitudinal ultrasonic assisted technology could reduce the bonding temperature at most by 180°C and bonding duration by 3-5s with ultrasonic parameter of 26W/cm 2 sound intensity and 4~4.66MPa prebonding pressure. This study provides valuable insights for optimizing flip-chip bonding processes.

关键词： Vibrations Temperature distribution Ultrasonic imaging Acoustics Flip-chip devices Bonding Surface cracks

来源：评论

学校读者我要写书评

暂无评论

Channel Attention Network for Feature Fusion of Ultrasonic and IMU Signals in Cross-Subject Gait Recognition

Channel Attention Network for Feature Fusion of Ultrasonic a...

引用

IEEE International Conference on systems, Man and Cybernetics

作者： Xujia Huang Zixiang Zhou Xinglong Xu Zhiyong Wang State Key Laboratory of Robotics and System Harbin Institute of Technology Shenzhen Shenzhen China

ISBN: (数字)9781665410205

ISBN: (纸本)9781665410212

In traditional human-machine interface research, a single sensing mode is often adopted, focusing on individual human motion classification experiments. However, the performance often declines when applied across individuals. Ultrasonic and inertial sensing, as mainstream technologies in human-machine interfaces, can respectively detect the muscle states and limb information of the human body, offering potential complementary benefits. The fusion of ultrasonic(US) and Inertial Measurement Unit(IMU) is expected to significantly enhance the generalization ability of physiological signals in motion intention recognition. This paper presents a method that employs fusion of ultrasound and IMU signals, and designs a Convolutional Neural Network with Channel Attention Module (IU-CAM-CNN) for gait classification experiments. The experimental results demonstrate that by introducing channel attention modules, the model's ability to extract features from fused signals can be effectively enhanced, resulting in a classification accuracy of 90% for cross-individual movements. In contrast, traditional LDA and CNN networks achieve only 22% and 47% classification accuracy, respectively. This method provides an effective solution for cross-individual movement intention recognition.

关键词： Ultrasonic imaging Accuracy Ultrasonic variables measurement Human-machine systems Feature extraction Physiology Acoustics Robustness Sensors Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Dual Transducers Co-Focusing Method for Ultrasound Stimulation in Calf Peripheral Nervous system: A Feasibility Study

Dual Transducers Co-Focusing Method for Ultrasound Stimulati...

引用

IEEE International Conference on systems, Man and Cybernetics

作者： Haoran Zheng Xiaoxin Wang Hongwei Zhang Honghai Liu State Key Laboratory of Robotics and System Harbin Institute of Technology Shenzhen Shenzhen China

ISBN: (数字)9781665410205

ISBN: (纸本)9781665410212

As a neuromodulation technology, ultrasound stimulation exhibits great potential due to its non-invasive and targeted nature. Low-Intensity Focused Ultrasound (LIFU) allows for precise targeting of the peripheral nervous system, enabling the elicitation of diverse sensations in the human body, including tactile, cold, heat, and pain. Current research efforts predominantly focus on ultrasound stimulation of the upper limbs, specifically the fingers. However, there is a paucity of investigation pertaining to the stimulation of the peripheral nervous system in the lower limb of the human body. Furthermore, the utilization of a singularly focused transducer for stimulation introduces anisotropy, thereby rendering it challenging to enhance the axial resolution. This study aims to explore an innovative approach using dual transducers that synergistically focus ultrasound to achieve highly precise stimulation and reduce the disparity between axial and lateral resolution. The primary objective is to investigate the feasibility of utilizing this approach for deep muscle stimulation in the human calf muscles, thus delving into the realm of lower limb neuromodulation.

关键词： Peripheral nervous system Legged locomotion Limbs Transducers Ultrasonic imaging Pain Muscles Brain modeling Rendering (computer graphics) Neuromodulation

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：