检索结果-内蒙古大学图书馆

13th ieee data driven control and learning systems conference, DDCLS 2024

作者： Wang, Qiang Huang, Xiao Tao, Hao Xia, Zhiwei Liu, Chun Ren, Xiaoqiang China Ship Development and Design Center Wuhan430064 China School of Mechatronic Engineering and Automation Shanghai University Shanghai200444 China Institute of Artificial Intelligence Shanghai University Shanghai200444 China

ISBN: (纸本)9798350361674

To address the challenges associated with marine navigation radar tracking in scenarios characterized by dense clutter and high-speed targets, this paper presents a multi-target tracking approach featuring an adaptive innovation sequence-based non-linear joint probability data association (AIS-JPDA) algorithm. The introduction of a straightforward calculation method for association probability aims to mitigate the issue of incorrect associations in target tracking within dense clutter environments. This method incorporates common measurements and their membership degrees to rectify the association probability, thereby enhancing the accuracy and real-time efficiency of multi-target track associations. The adaptive innovation policy is integrated to allow for the dynamic adjustment of the covariance matrix, accommodating process and measurement noise in non-linear maneuvering models. Furthermore, a lifecycle-based target track maintenance method is introduced to tackle the challenge of track disconnection resulting from high target relative velocity. Finally, simulation results demonstrate the effectiveness of the proposed algorithm. © 2024 ieee.

关键词： Covariance matrix

来源：评论

学校读者我要写书评

暂无评论

Tiny Machine learning for Li-ion Battery State of Health Estimation 22

Tiny Machine Learning for Li-ion Battery State of Health Est...

引用

ieee 22nd Mediterranean Electrotechnical conference (MELECON)

作者： Giazitzis, Spyridon Sakwa, Maciej Ogliari, Emanuele Badha, Susheel Rosetti, Filippo Politecn Milan Dept Energy Milan Italy Infineon Technol Villach Austria

ISBN: (纸本)9798350387032;9798350387025

The high rise of Renewable Energy Sources (RES) and the stochastic nature of weather conditions cause a need for battery integration into energy storage systems. Sophisticated Battery Management systems (BMS) are required to supervise and compute the key parameters among which many are not directly measurable using standard sensors. Many data-driven methods are frequently preferred for their estimation due to their superior accuracy, minimal energy consumption, and straightforward implementation on cloud platforms or Internet of Things (IoT) devices. The primary focus of this research is the development of compact, data-driven models tailored for battery State of Health (SoH) estimation prioritizing efficiency in terms of memory and power requirements, rendering them well-suited for integration into IoT devices. Specifically, the target device for implementation is the Infineon Cypress CY8CPROTO-062S3-4343W. The development process entailed offline training of a compact Convolutional Neural Network (CNN) and an Artificial Neural Network (ANN) utilizing a comprehensive dataset derived from five distinct batteries. The on-device deployment and quantization of the models was performed employing the ModusToolBox Machine learning (MTB-ML) software provided by Infineon. The tests demonstrate satisfactory results for both selected models, with notably good accuracy, particularly in the initial phases of the battery life cycle. Regarding computational load, the ANN exhibits a distinct advantage over the more intricate CNN model.

关键词： State of Health TinyML Battery Deep learning IoT

来源：评论

学校读者我要写书评

暂无评论

A data-driven Approach to Geometric Modeling of systems with Low-Bandwidth Actuator Dynamics

A Data-Driven Approach to Geometric Modeling of Systems with...

引用

ieee International conference on Robotics and Automation (ICRA)

作者： Deng, Siming Liu, Junning Datta, Bibekananda Pantula, Aishwarya Gracias, David H. Nguyen, Thao D. Bittner, Brian A. Cowan, Noah J. Johns Hopkins Univ Lab Computat Sensing & Robot Baltimore MD 21218 USA Johns Hopkins Univ Dept Mech Engn Baltimore MD 21218 USA Johns Hopkins Univ Dept Chem & Biomol Engn Baltimore MD 21218 USA Johns Hopkins Univ Appl Phys Lab Laurel MD 97331 USA

ISBN: (纸本)9798350384581;9798350384574

It is challenging to perform system identification on soft robots due to their underactuated, high-dimensional dynamics. In this work, we present a data-driven modeling framework, based on geometric mechanics (also known as gauge theory) that can be applied to systems with low-bandwidth control of the system's internal configuration. This method constructs a series of connected models comprising actuator and locomotor dynamics based on data points from stochastically perturbed, repeated behaviors. By deriving these connected models from general formulations of dissipative Lagrangian systems with symmetry, we offer a method that can be applied broadly to robots with first-order, low-pass actuator dynamics, including swelling-driven actuators used in hydrogel crawlers. These models accurately capture the dynamics of the system shape and body movements of a simplified swimming robot model. We further apply our approach to a stimulus-responsive hydrogel simulator that captures the complexity of chemomechanical interactions that drive shape changes in biomedically relevant micromachines. Finally, we propose an approach of numerically optimizing control signals by iteratively refining models, which is applied to optimize the input waveform for the hydrogel crawler. This transfer to realistic environments provides promise for applications in locomotor design and biomedical engineering.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Machine learning Techniques for Task Offloading in Connected Autonomous Vehicles 3

Machine Learning Techniques for Task Offloading in Connected...

引用

3rd ieee International conference on Power Electronics, Intelligent control and Energy systems, ICPEICES 2024

作者： Kumar, Bhoopendra Bhardwaj, Aditya Sahu, Dinesh School of Scset Bennett University Greater Noida India

ISBN: (纸本)9798350387193

Connected Autonomous Vehicles (CAVs) can revolutionize the haulage industry by improving safety, efficiency, and comfort. However, real-time data processing within latency constraints is crucial for autonomous vehicles to perform well. Due to the limited computational resources available in CAVs, it is important to utilize them efficiently. One way to address this challenge is task offloading, which involves delegating computationally intensive tasks to remote servers or edge devices. In this context, a machine learning-based task offloading system is proposed for CAV s' edge computing networks. This system would dynamically determine the optimal offloading decisions regarding the vehicle's context & network conditions. Additionally, the paper provides a critical review and future roadmap for ML-driven task offloading in CAVs' edge computing environments. Finally, future directions for task offloading in the CAV s system are explored to provide valuable insights for researchers seeking to contribute to advancing the task offloading system. © 2024 ieee.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Recognition of Scalp Psoriasis Dermoscopic Patterns Signs Based on Histogram Equalization and Convolutional Neural Networks 13

Recognition of Scalp Psoriasis Dermoscopic Patterns Signs Ba...

引用

13th ieee data driven control and learning systems conference, DDCLS 2024

作者： Ji, Honghai Li, Jiaqi Jiang, Weiwei Wang, Li Liu, Shida Chen, Yang School of Electronics & Control Engineering North China University of Technology Beijing100144 China 72nd Group Army Hospital of Pla Department of Dermatology Zhejiang313000 China

ISBN: (纸本)9798350361674

Scalp psoriasis is a common chronic autoimmune inflammatory skin disease that affects a large number of people worldwide. Although scalp psoriasis is not physically a serious life-threatening condition for patients, the itchiness, inflammation, and social distress it causes have a significant negative impact on their quality of life. Currently, the urgent need for accurate diagnosis of scalp psoriasis has driven researchers to explore efficient diagnostic techniques utilizing deep learning techniques, among which dermoscopy is widely used as a simple and easy-to-use non-invasive testing device. In this paper, a convolutional neural network (HE-CNN) based on histogram equalization is proposed to solve the problem of automatic recognition of different dermoscopic patterns of scalp psoriasis. In this, histogram equalization is used to perform global brightness enhancement of dermoscopic images to ensure that important lesion features in the image are not obscured by hair shadows leading to neglect. CBAM is added to the CNN and the enhanced images are used to train a classification model for automated recognition of scalp psoriasis for clinical detection. We conducted experiments on a scalp psoriasis dataset provided by 72nd Group army hospital of PLA. In order to verify the effectiveness of the method, we use the unenhanced image as a control group and compare it with the image that has undergone enhancement, and the results after five-fold cross-validation show that the recognition effect of HE-CNN outperforms that of the control group in terms of accuracy, specificity, sensitivity and youden index. It is verified that HE-CNN has better effect in assisting doctors in the diagnosis of scalp psoriasis. © 2024 ieee.

关键词： Equalizers

来源：评论

学校读者我要写书评

暂无评论

Deep Reinforcement learning for Intelligent Transportation systems: A Survey

引用

ieee TRANSACTIONS ON INTELLIGENT TRANSPORTATION systems 2022年第1期23卷 11-32页

作者： Haydari, Ammar Yilmaz, Yasin Univ S Florida Dept Elect Engn Tampa FL 33620 USA Univ Calif Davis Dept Elect & Comp Engn Davis CA 95616 USA

Latest technological improvements increased the quality of transportation. New data-driven approaches bring out a new research direction for all control-based systems, e.g., in transportation, robotics, IoT and power systems. Combining data-driven applications with transportation systems plays a key role in recent transportation applications. In this paper, the latest deep reinforcement learning (RL) based traffic control applications are surveyed. Specifically, traffic signal control (TSC) applications based on (deep) RL, which have been studied extensively in the literature, are discussed in detail. Different problem formulations, RL parameters, and simulation environments for TSC are discussed comprehensively. In the literature, there are also several autonomous driving applications studied with deep RL models. Our survey extensively summarizes existing works in this field by categorizing them with respect to application types, control models and studied algorithms. In the end, we discuss the challenges and open questions regarding deep RL-based transportation applications.

关键词： Machine learning Autonomous vehicles control systems Markov processes Deep reinforcement learning intelligent transportation systems traffic signal control autonomous driving multi-agent systems

来源：评论

学校读者我要写书评

暂无评论

Comprehensive data-driven Framework for Detecting and Classifying Non-Technical Distribution Losses

引用

ieee ACCESS 2024年 12卷 96537-96551页

作者： Jene-Vinuesa, Marc Aragues-Penalba, Monica Sumper, Andreas Univ Politecn Cataluna Ctr Innovacio Tecnol Convertidors Estat Accionamen Dept Engn Elactr ETS Engn Ind Barcelona Barcelona 08028 Spain

Non-technical losses (NTL) pose multiple challenges across distribution grids. This paper introduces a comprehensive framework combining data-driven methods to detect and categorize fraud due to meter tampering or direct connections while identifying potential culprits. The hybrid methodology utilizes grid and consumer-related data to obtain NTL curves through an energy balance approach, yielding indicators such as magnitude, duration, and other features. A Random Forest classifier trained with real historical cases of NTL achieves a weighted F1 score of 0.859, effectively labelling fraud types. Additionally, an unsupervised detection model, integrating clustering and correlation methods, enables accurate identification of tampered meters. The paper introduces two adjustable parameters enabling utilities to fine-tune meter tampering detection strategies based on economic considerations. The results demonstrate that true positives can be increased at the expense of increasing false positives. Accurate fraud identification is achieved using the Fuzzy C-Means algorithm, with an F1 score of 0.9. The algorithm is tested on grids with distributed generation, with a decrease of 10% on the predictive performance when half of the users are prosumers, demonstrating the methodology's promising performance in real-world scenarios.

关键词： Fraud Classification algorithms Meters Voltage control Support vector machines Hardware Clustering algorithms Unsupervised learning Classification direct connection false data injection (FDI) fraud detection hybrid algorithm inspection strategies meter tampering non-technical losses (NTL) unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Meta-learning of data-driven controllers with automatic model reference tuning: theory and experimental case study 63

Meta-learning of data-driven controllers with automatic mode...

引用

63rd ieee conference on Decision and control, CDC 2024

作者： Busetto, Riccardo Breschi, Valentina Baracchi, Federica Formentin, Simone Politecnico di Milano Dip. di Elettronica Informazione e Bioingegneria Milano Italy Eindhoven University of Technology Dept. of Electrical Engineering Eindhoven Netherlands

ISBN: (纸本)9798350316339

data-driven control offers a viable option for control scenarios where constructing a system model is expensive or time-consuming. Nonetheless, many of these algorithms are not entirely automated, often necessitating the adjustment of multiple hyperparameters through trial-and-error processes and demanding significant amounts of data. In this work, we explore a meta-learning approach to leverage prior knowledge about analogous (though not identical) systems, aiming to reduce both the experimental workload and ease the tuning of the available degrees of freedom. We validate this methodology through an experimental case study involving the tuning of proportional, integral (PI) controllers for brushless DC (BLDC) motors with variable loads and architectures. © 2024 ieee.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learned Lifted Linearization Applied to Unstable Dynamic systems Enabled by Koopman Direct Encoding

引用

ieee control systems LETTERS 2023年 7卷 1153-1158页

作者： Ng, Jerry Asada, H. Harry MIT Dept Mech Engn Cambridge MA 02139 USA

This letter presents a Koopman lifting linearization method that is applicable to nonlinear dynamical systems having both stable and unstable regions. It is known that Dynamic Mode Decomposition (DMD) and its extended methods are often unable to model unstable systems accurately and reliably. Here we solve the problem through merging three methodologies: decomposition of a lifted linear system into stable and unstable modes, deep learning of a dictionary of observable functions in the separated subspaces, and a new formula for obtaining the Koopman operator, called Direct Encoding. Two sets of effective observable functions are obtained through neural net training where the training data are separated into stable and unstable trajectories. The resultant learned observables are used for lifting the state space, and a linear state transition matrix is constructed using Direct Encoding where inner products of the learned observables are computed. The proposed method shows a dramatic improvement over existing DMD and data-driven methods. Furthermore, a method is developed for determining the boundaries between stable and unstable regions.

关键词： Neural networks modeling machine learning stability of nonlinear systems

来源：评论

学校读者我要写书评

暂无评论

Improved quantized iterative learning control in the presence of data dropouts 14

Improved quantized iterative learning control in the presenc...

引用

14th Asian control conference (ASCC)

作者： Shiao, Zhen Chen, Yu Kong, Fanrong Hohai Univ Coll Artificial Intelligence & Automat Nanjing 211100 Peoples R China

ISBN: (纸本)9798331540845;9789887581598

This paper focuses on solving the simultaneous data quantization and data dropouts problems in multi-input multi-output unknown nonlinear discrete systems, with an emphasis on their full tracking control. First, an innovative quantization model-free adaptive iterative learning control scheme is proposed based on the dynamic linearization model of the system. Using a uniform quantizer with an improved encoding-decoding mechanism to quantise the system output, modelling the data dropouts problem as a Bernoulli sequence, and employing intermittent updating to deal with the effects of the random data dropouts problem in the system output, the inherent errors due to data quantization and data dropouts are successfully eliminated, and it is verified that only a small amount of input/output data is required to achieve the tracking and controlling error convergence. Finally, the effectiveness of the method can be demonstrated through theoretical derivation and simulation experiments.

关键词： data quantization data dropouts Quantitative model-free adaptive iterative learning control

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：