检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Li, Xibo Patel, Shruti Stronzek-Pfeifer, David Büskens, Christof The Working Group Optimization and Optimal Control Center for Industrial Mathematics University of Bremen Bremen28359 Germany

Global navigation satellite systems readily provide accurate position information when localizing a robot outdoors. However, an analogous standard solution does not exist yet for mobile robots operating indoors. This paper presents an integrated framework for indoor localization and experimental validation of an autonomous driving system based on an advanced driver-assistance system (ADAS) model car. The global pose of the model car is obtained by fusing information from fiducial markers, inertial sensors and wheel odometry. In order to achieve robust localization, we investigate and compare two extensions to the Extended Kalman Filter;first with adaptive noise tuning and second with χ2−testing for measurement outlier detection. An efficient and low-cost ground truth measurement method using a single LiDAR sensor is also proposed to validate the results. The performance of the localization algorithms is tested on a complete autonomous driving system with trajectory planning and model predictive control. Copyright © 2023, The Authors. All rights reserved.

关键词： Fault detection

来源：评论

学校读者我要写书评

暂无评论

Unbalanced few-shot classification based on ME-MAML framework

Unbalanced few-shot classification based on ME-MAML framewor...

引用

第43届中国控制会议

作者： Zherun Liao Zheng Lv Qiao Sun Jun Zhao Key Laboratory of Intelligent Control and Optimization for Industrial Equipment of Ministry of Education Dalian University of Technology Electric Operations Control Center State Grid Dalian Power Supply Company

ISBN: (数字)9789887581581

ISBN: (纸本)9798350366907

Aiming at the few-shot problem of lack of sufficient samples and uneven samples, which is common in medical image classification tasks, this paper proposes a meta-learning framework ME-MAML(Medical ModelAgnostic Meta-Learning) for few-shot medical image classification tasks. The framework aims to train a set of parameters with good generalization performance for the multi-class medical image classification task by learning the base class of medical images, and complete the generalization for the new class of medical image classification task by learning a small number of new classes of medical images. ME-MAML improves the MAML(ModelAgnostic Meta-Learning) meta-learning framework from the aspects of the feature extraction network, the loss function and the algorithmic framework, which makes the framework more suitable for few-shot medical image classification tasks. The experimental results demonstrate that for few-shot medical image classification task, the framework has better classification performance compared to supervised learning frameworks as well as advanced meta-learning frameworks.

关键词： Few-shot Learning Meta-learning Medical Image Classification

来源：评论

学校读者我要写书评

暂无评论

Investigating the Power Budget of a 3U Nanosatellite Designed for Earth Observation 10

Investigating the Power Budget of a 3U Nanosatellite Designe...

引用

10th IEEE International Workshop on Metrology for AeroSpace, MetroAeroSpace 2023

作者： Daghouri, Amina El Hachimi, Youssef Ouhammam, Abdelaali Chanoui, Mohammed Alae Elhani, Soumia Mahmoudi, Hassane ENSAM University Center for Research in Space Technologies Energy Optimization Diagnosis and Control Team EMI Mohammed v University Rabat Morocco Power Electronics Smart Control Techniques Automatic and Robotics Team EMI University Center for Research in Space Technologies EMI Mohammed v University Rabat Morocco EST University Center for Research in Space Technologies LASTIMI Laboratory EMI Mohammed v University Rabat Morocco

ISBN: (纸本)9781665456906

The Electrical Power System (EPS) is one of the most critical subsystems of a nanosatellite. Thus, its design should be done carefully for a successful mission. To verify that the power needs of its subsystems does not exceed the power available, a power budget should be established. The purpose of this work is to evaluate the 3U nanosatellite's EPS by analyzing its power budget using Systems Tool Kit (STK) software. In this study, the orbital design, the sizing of the PV panels and the batteries used are presented. The simulation is done taking into account the worst case assumptions: the sun-synchronous Low Earth Orbit (LEO) orbit with the longest eclipse period, the worst case solar irradiance of 1316 W/m2. The results of the power budget established confirm the selection and the sizing of PV panels and the batteries. © 2023 IEEE.

关键词： Electric power utilization

来源：评论

学校读者我要写书评

暂无评论

Let Hybrid A* Path Planner Obey Traffic Rules: A Deep Reinforcement Learning-Based Planning Framework

Let Hybrid A* Path Planner Obey Traffic Rules: A Deep Reinfo...

引用

International Conference on Autonomous Robots and Agents, ICARA

作者： Xibo Li Shruti Patel Christof Büskens Optimization and Optimal Control Center for Industrial Mathematics University of Bremen Bremen Germany Department for Autonomous Systems TOPAS Industriemathematik Innovation gGmbH Bremen Germany

ISBN: (数字)9798331509231

ISBN: (纸本)9798331509248

Deep reinforcement learning (DRL) allows a system to interact with its environment and take actions by training an efficient policy that maximizes self-defined rewards. In autonomous driving, it can be used as a strategy for high-level decision making, whereas low-level algorithms such as the hybrid A* path planning have proven their ability to solve the local trajectory planning problem. In this work, we combine these two methods where the DRL makes high-level decisions such as lane change commands. After obtaining the lane change command, the hybrid A* planner is able to generate a collision-free trajectory to be executed by a model predictive controller (MPC). In addition, the DRL algorithm is able to keep the lane change command consistent within a chosen time-period. Traffic rules are implemented using linear temporal logic (LTL), which is then utilized as a reward function in DRL. Furthermore, we validate the proposed method on a real system to demonstrate its feasibility from simulation to implementation on real hardware.

关键词： Training Trajectory planning Predictive models Deep reinforcement learning Prediction algorithms Planning Trajectory Logic Autonomous vehicles Robots

来源：评论

学校读者我要写书评

暂无评论

Analysis of CubeSat thermal performance using various PV panel configurations 10

Analysis of CubeSat thermal performance using various PV pan...

引用

10th IEEE International Workshop on Metrology for AeroSpace, MetroAeroSpace 2023

作者： Narimane, Blanchete Daghouri, Amina Bah, Abdellah El Hani, Soumia ENSAM University Center for Research in Space Technologies EMI Mohammed v University Royal Centre for Space Research and Studies Thermal and Energy Research Team Rabat Morocco ENSAM University Center for Research in Space Technologies Energy Optimization Diagnosis and Control Team EMI Mohammed v University Rabat Morocco

ISBN: (纸本)9781665456906

Managing the temperature of a nanosatellite is crucial to ensure the proper functioning of the payload and the platform during a mission. A thermal engineer faces the challenge of keeping both within acceptable temperature margins, which can be affected by various factors such as solar radiation, heat generated by on board equipments, and the spacecraft's orientation. To address this challenge, a simplified thermal model is typically created and examined. This model takes into account the spacecraft's geometry, material properties, and environmental factors to determine the temperature of critical components. This paper presents the results of simulations conducted on a 3D CubeSat orbiting at an altitude of 600Km. The simulations explore various PV panel combinations to determine the best compromise between thermal control and allowable power. This involves finding the optimal PV panel configuration that provides the necessary thermal control while still allowing the spacecraft to operate within its power constraints. © 2023 IEEE.

关键词： Thermoanalysis

来源：评论

学校读者我要写书评

暂无评论

DESIGN OF EUTROPHICATION MONITORING PLATFORM FOR CHAOHU CONNECTED RIVER CHANNEL BASED ON LoRa AND DEEP LEARNING

UPB Scientific Bulletin, Series C: Electrical Engineering an...

引用

UPB Scientific Bulletin, Series C: Electrical Engineering and Computer Science 2023年第3期85卷 285-296页

作者： Kong, Bing Yu, Mei Ye, Song Xu, Xiaoyong Xiang, Rong School of Electronic Engineering Chaohu University Hefei Anhui China Graduate School Angeles University Foundation Manila Philippines School of Chemistry & Material Engineering Chaohu University Hefei Anhui China Industrial Process Control Optimization and Automation Engineering Research Center Chaohu University Hefei Anhui China Key Laboratory of Novel Ceramic and Powder Engineering Chaohu University Anhui China

Traditional methods cannot meet the requirements of real-time monitoring and scientific prevention and control of blue-green algae bloom. This paper uses multi-parameter sensors, LoRa module, 4G module, PLC and shore-based video acquisition device to build a monitoring platform to achieve real-time monitoring of multiple locations and real-time collection of video images;Aiming at the problem of image motion blur caused by shore-based video in water surface fluctuation, the Multi-scale Revolution Neural Network is used to effectively eliminate the visual artifacts of motion blur image, improve the monitoring accuracy of blue-green algae bloom, and meet the actual needs. © 2023, Politechnica University of Bucharest. All rights reserved.

关键词： Eutrophication

来源：评论

学校读者我要写书评

暂无评论

Research on Blood Glucose Regulation of Type I Diabetes Based on Adaptive RBF Network

Research on Blood Glucose Regulation of Type I Diabetes Base...

引用

2022 Chinese Automation Congress, CAC 2022

作者： Liu, YunShuan Sun, KaiBiao Dalian University of Technology The Key Laboratory of Intelligent Control and Optimization for Industrial Equipment Ministry of Education Liaoning Dalian China Dalian University of Technology Professional Technology Innovation Center Distributed Control for Industrial Equipment of Liaoning Province Dalian China

ISBN: (纸本)9781665465335

A backstepping control algorithm based on an adaptive RBF neural network is proposed for the glucose regulation problem in type I diabetes. Firstly, the complex mechanism of glucose-insulin action in human body is described based on Bergman minimum model;secondly, the RBF neural network are used to approximate the nonlinear functions in Bergman's mini-mum model, then the adaptive rate of the proposed algorithm is derived and the controller is designed based on the backstepping method to ensure the stability and convergence of the whole system;finally, the validation are performed by MATLAB and compared with the results of blood glucose control in recent years. The final results show that the controller has the advantages of fast convergence, insensitivity to external interference and small control quantity. © 2022 IEEE.

关键词： controllers

来源：评论

学校读者我要写书评

暂无评论

Let Hybrid A∗ Path Planner Obey Traffic Rules: A Deep Reinforcement Learning-Based Planning Framework

arXiv

引用

arXiv 2024年

作者： Li, Xibo Patel, Shruti Büskens, Christof The working group Optimization and Optimal Control Center for Industrial Mathematics University of Bremen Bremen28359 Germany TOPAS Industriemathematik InnovationgGmbH Bremen28359 Germany

Deep reinforcement learning (DRL) allows a system to interact with its environment and take actions by training an efficient policy that maximizes self-defined rewards. In autonomous driving, it can be used as a strategy for high-level decision making, whereas low-level algorithms such as the hybrid A∗ path planning have proven their ability to solve the local trajectory planning problem. In this work, we combine these two methods where the DRL makes high-level decisions such as lane change commands. After obtaining the lane change command, the hybrid A∗ planner is able to generate a collision-free trajectory to be executed by a model predictive controller (MPC). In addition, the DRL algorithm is able to keep the lane change command consistent within a chosen time-period. Traffic rules are implemented using linear temporal logic (LTL), which is then utilized as a reward function in DRL. Furthermore, we validate the proposed method on a real system to demonstrate its feasibility from simulation to implementation on real hardware. © 2024, CC BY.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

Zero-shot industrial image denoising with lightweight network

Zero-shot industrial image denoising with lightweight networ...

引用

2024 International Workshop on Automation, control, and Communication Engineering, IWACCE 2024

作者： Zhou, Chengxu Fan, Xiaojie Wang, Simeng Liu, Hongyan Gu, Ke School of Information Science and Technology Beijing University of Technology Beijing China Engineering Research Center of Intelligent Perception and Autonomous Control of Ministry of Education China Beijing Laboratory of Smart Environmental Protection China Beijing Artificial Intelligence Institute China School of Electronic & Information Engineering Liaoning University of Technology Liaoning China Key Laboratory of Intelligent Control and Optimization for Industrial Equipment of Ministry of Education Dalian University of Technology China

ISBN: (数字)9781510685420

ISBN: (纸本)9781510685413

Industrial images are often captured under full-time and full-weather conditions, leading to inevitable noise during the imaging process, which can impact subsequent detection algorithms. In recent years, image denoising with neural networks has been the rapid development. However, training such networks typically requires a large dataset, which is scarce in publicly available industrial image databases. In this paper, we propose a novel approach termed Zero-Shot Industrial Image Lightweight Denoising (ZSILD) network, which effectively denoises single noisy industrial image without the need for datasets. First, we sample the paired neighbour pixels of a random noisy industrial image, which are then utilized to train a lightweight denoising network. Second, we design a lightweight depthwise convolutions network based on bottleneck residual structure with shortcut connections. Finally, this network is trained on the sampled pairs using a novel loss function aimed at enhancing denoising performance. Our experiments conduct on real-world industrial ambient noise demonstrate that our ZSILD method outperforms existing denoising techniques, all while requiring comparatively minimal computational resources. © 2024 SPIE. All rights reserved.

关键词： Image denoising

来源：评论

学校读者我要写书评

暂无评论

Short-Term Reactive Load Forecasting Based on a Hybrid Machine Learning Model

Short-Term Reactive Load Forecasting Based on a Hybrid Machi...

引用

2022 Power System and Green Energy Conference, PSGEC 2022

作者： Qi, Le Chen, Mingyuan Song, Jifeng Zhang, Minyu Zheng, Wenbin Shang, Qinghua Power Dispatching Control Center of Guangxi Power Grid Co. Ltd Nanning China Guangxi University Key Laboratory of Guangxi Electric Power System Optimization and Energy-Savig Technology Nanning China

ISBN: (数字)9781665499934

ISBN: (纸本)9781665499934

With the improvement of people's living standards and the access to new energy sources such as rooftop photovoltaics, the variety of electricity consumption on the customer side has increased dramatically, placing high demands on the reactive power compensation configuration, regulation, and control capabilities at all levels of the distribution network. Predicting reactive loads directly impacts the quality of power supply and is also an integral part of reactive power optimization in the grid. With the popularity of smart energy measurement devices, large-scale historical reactive load data is uploaded to the dispatch control system. This paper proposes a hybrid reactive load prediction model based on a machine learning framework. Appropriate error parameters are selected, and the model is compared with a traditional time series prediction model to verify that the model can achieve better prediction results. © 2022 IEEE.

关键词： Machine learning

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：