检索结果-内蒙古大学图书馆

ieee 12th data driven control and learning systems conference (DDCLS)

作者： Zhang, Yanyan Zhang, Kai Yang, Pengcheng Peng, Kaixiang Univ Sci & Technol Beijing Sch Automat & Elect Engn Beijing 100083 Peoples R China Minist Educ Key Lab Knowledge Automat Ind Proc Beijing 100083 Peoples R China

ISBN: (纸本)9798350321050

Due to the difficulty in strip crown prediction caused by multivariable, nonlinear and strong coupling in the hot strip rolling mill (HSRM) process, this paper proposes a strip crown prediction model based on support vector regression (SVR), and uses sparrow search algorithm (SSA) to optimize the parameter C and. of the model, so as to improve the generalization ability of the prediction model. The overall performance of the model is evaluated by mean square error (MSE), mean absolute error (MAE), mean absolute percentage error (MAPE) and correlation coefficient ( R2). It shows that the prediction accuracy and generalization ability of the proposed model are better than the traditional methods. The proposed SSA-SVR model in this paper is successfully applied to the crown prediction of the 2150 production line of Ansteel company. The performance shows that the method can be efficient to predict the steel crown in a real HSRM process.

关键词： Strip crown hot strip rolling mill process support vector regression sparrow search algorithm prediction

来源：评论

学校读者我要写书评

暂无评论

Iteration Number Bound for Adaptive learning control of Nonlinear systems under Specified Approximation Error 13

Iteration Number Bound for Adaptive Learning Control of Nonl...

引用

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

作者： Huang, Dunsheng Shen, Dong School of Mathematics Renmin University of China Beijing100872 China

ISBN: (纸本)9798350361674

This study addresses the tracking problem for a class of nonlinear uncertain systems over a finite time interval, where the unknown and nonparametric nature of the system's nonlinear functions is taken into consideration. In the context of adaptive iterative learning control, it becomes necessary to employ known basis functions in order to approximate the nonlinear uncertainty for parameterizations. These basis functions must be sufficient to ensure convergence of tracking error. The relationship between the number of bases used and the iteration count plays a significant role in achieving desired tracking accuracy. Additionally, when applying adaptive iterative learning control, resetting the initial state of the system introduces uncertainty in terms of initial state error, which affects the iteration count. In this study, we examine five different conditions related to initial state error and establish an upper bound on iterations that is proportional to approximation error under these conditions. Our analysis provides valuable insights for accelerating convergence rate. © 2024 ieee.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

data-driven Tracking control for Multi-Agent systems With Unknown Dynamics via Multithreading Iterative Q-learning

引用

ieee TRANSACTIONS ON systems MAN CYBERNETICS-systems 2023年第4期53卷 2533-2542页

作者： Dong, Tao Gong, Xiaomei Wang, Aijuan Li, Huaqing Huang, Tingwen Southwest Univ Coll Elect & Informat Engn Chongqing 400715 Peoples R China Chongqing Univ Technol Coll Comp Sci & Engn Chongqing 400054 Peoples R China Texas A&M Univ Qatar Dept Math Doha Qatar

This article addresses the tracking control problem of multiagent systems (MASs) with unknown dynamics. First, by designing a compensator, an augmented neighborhood error system is proposed. Then, a multithreading iterative $Q$ -learning algorithm is developed, which can transform the tracking control problem into the optimal regulation of the augmented error system. In this algorithm, the agent controller is multithread, which consists of thread unit, global unit, and replay buffer. The thread unit is used to create the thread to collect the transitions with different control policies, which can reduce data relativity and improve data collection speed and exploration capability. The replay buffer is used to store the generated transitions. The global unit is used to generate the control policy. Moreover, the convergence analysis of the proposed algorithm is given, which shows that the neighborhood error converges to zero and the tracking control problem can be solved. Finally, the numerical example not only verifies the effectiveness of the obtained results but also shows the proposed algorithm performs better than the $Q$ -iterative algorithm.

关键词： Iterative Q-learning multiagent systems (MASs) multithreading tracking control

来源：评论

学校读者我要写书评

暂无评论

Self-learning Temporal-Spatial Graph Model for Industrial Soft Sensing Application 12

Self-Learning Temporal-Spatial Graph Model for Industrial So...

引用

ieee 12th data driven control and learning systems conference (DDCLS)

作者： Zhang, Chiye Yao, Le Chen, Zhichao Shen, Bingbing Zeng, Jiusun Zhejiang Univ State Key Lab Ind Control Technol Coll Control Sci & Engn Hangzhou 310027 Peoples R China Hangzhou Normal Univ Sch Math Hangzhou 311121 Peoples R China

ISBN: (纸本)9798350321050

In recent years, deep learning (DL) has been widely studied in soft sensing due to its advantages in feature extraction and nonlinear fitting. However, DL models are black-box models, which are difficult to explain. It is difficult to introduce the prior knowledge about the process into the model, and these shortcomings hinder its application in the actual industrial processes. In this paper, a Self-learning Temporal-Spatial Graph (SLTSG) deep learning model is proposed and applied to the soft sensor modeling of complex dynamic industrial processes. It uses the graph attention network to construct the explicit nonlinear relationship between variables, and uses its ability to aggregate information to complete the feature extraction from the graph for each time step. Then, the spatio-temporal convolutional network is utilized to extract local nonlinear dynamic features for the prediction of quality variables. Since it is difficult and subjective to obtain graph structure by relying on prior knowledge such as process mechanism, a self-learning method of graph structure is adopted in this paper. To verify the effectiveness of the proposed method, the soft sensing case study of carbon monoxide content in the process of High-Low Transformer is presented. The experimental results show that the model not only has high accuracy, but also has certain knowledge discovery ability.

关键词： Soft Sensing Graph Attention Mechanism Deep learning Graph Structure learning Graph Neural Network

来源：评论

学校读者我要写书评

暂无评论

Augmented Industrial data-driven Modeling Under the Curse of Dimensionality

引用

ieee/CAA Journal of Automatica Sinica 2023年第6期10卷 1445-1461页

作者： Xiaoyu Jiang Xiangyin Kong Zhiqiang Ge the State Key Laboratory of Industrial Control Technology College of Control Science and EngineeringZhejiang UniversityHangzhou 310027China the Peng Cheng Laboratory Shenzhen 518000China

The curse of dimensionality refers to the problem o increased sparsity and computational complexity when dealing with high-dimensional *** recent years,the types and vari ables of industrial data have increased significantly,making data driven models more challenging to *** address this prob lem,data augmentation technology has been introduced as an effective tool to solve the sparsity problem of high-dimensiona industrial *** paper systematically explores and discusses the necessity,feasibility,and effectiveness of augmented indus trial data-driven modeling in the context of the curse of dimen sionality and virtual big ***,the process of data augmen tation modeling is analyzed,and the concept of data boosting augmentation is *** data boosting augmentation involves designing the reliability weight and actual-virtual weigh functions,and developing a double weighted partial least squares model to optimize the three stages of data generation,data fusion and *** approach significantly improves the inter pretability,effectiveness,and practicality of data augmentation in the industrial ***,the proposed method is verified using practical examples of fault diagnosis systems and virtua measurement systems in the *** results demonstrate the effectiveness of the proposed approach in improving the accu racy and robustness of data-driven models,making them more suitable for real-world industrial applications.

关键词： Index Terms—Curse of dimensionality data augmentation data-driven modeling industrial processes machine learning

来源：评论

学校读者我要写书评

暂无评论

Human-Inspired Anticipative Cruise control for Enhancing Mixed Traffic Flow

引用

ieee TRANSACTIONS ON INTELLIGENT TRANSPORTATION systems 2024年第11期25卷 17335-17351页

作者： Nie, Zifei Farzaneh, Hooman Kyushu Univ Interdisciplinary Grad Sch Engn Sci Fukuoka 8168580 Japan

Introducing the human driver's intelligence of interpreting and anticipating preceding traffic conditions into automated driving control is crucial for improved driving experience. Existing human-like automated driving systems imitate human driver's reactive control patterns without reflecting anticipative behaviors. These systems involve either online solving predictive optimization or offline training machine learning techniques, which are computationally expensive and impractical for current production vehicles. Additionally, these systems are less effective in traffic environments with lean vehicular connectivity. This paper proposes a practical human-inspired anticipative cruise control for connected automated vehicles (CAVs) that not only respects human drivers' reactive driving preferences but exploits their anticipative capabilities to enhance mixed traffic flow. Two major contributions are (i) learning the reactive control objective from a segment of the human driving trajectory based on a recovery matrix;and (ii) data-enabled preceding traffic anticipation leveraging the behavioral theory of linear systems, which provides the estimated motion of immediate preceding vehicle (IPV) at a personalized anticipation horizon, considering the downstream traffic trends from a distant connected human-driven vehicle (CHV). The proposed control strategy is evaluated on a human-in-the-loop (HiL) driving simulator. Results validate its real-time implementation potential in (a) learning human-like reactive driving control;(b) improving traffic flow energy economy by 4.58%;(c) enhancing traffic efficiency by attenuating latent traffic fluctuations. Moreover, result reveals that the capability of mitigating traffic fluctuations is affected by the anticipation horizon, which can be adaptably learned from human driver's timely guidance by virtue of their intelligence of adaptability to dynamic traffic conditions.

关键词： Vehicles Behavioral sciences Cruise control Psychology Trajectory Training Market research Connected vehicles human-like driving control mixed traffic data-enabled traffic anticipation objective learning

来源：评论

学校读者我要写书评

暂无评论

A Salp Swarm Algorithm-based Model-free Adaptive control Method and Its Application in Transformer Constant Temperature System 12

A Salp Swarm Algorithm-based Model-free Adaptive Control Met...

引用

ieee 12th data driven control and learning systems conference (DDCLS)

作者： Li, Zhen Jin, Lei Ma, Guangjuan Liu, Shida North China Univ Technol Sch Elect & Control Engn Beijing 100093 Peoples R China Tbea Tianjin Transformers Co Ltd Tianjin 300308 Peoples R China

ISBN: (纸本)9798350321050

Aiming at the problem of temperature control of transformer, a salp swarm algorithm-based model free adaptive control (SSA-MFAC) algorithm is proposed. In SSA-MFAC, the nonlinear dynamics of the transformer temperature control system is linearized using the dynamic linearization technology and a time-varying parameters named pseudo-partial derivative (PPD). Then, an SSA-MFAC controller is designed, and the salp swarm algorithm (SSA) is introduced to automatically optimize the controller parameters. Based on the salps group behavior in the ocean, the SSA method achieves the goal of controller parameter optimization by simulating the predatory behavior of the salps. The main advantage of SSA-MFAC method is that the controller only makes use of the operation data of the transformer constant temperature oil bath refrigeration unit instead of the specific mathematical model, and the controller parameters are obtained by using the SSA algorithm instead of experience. The algorithm is proved to be effective based on MATLAB/Simulink simulation platform.

关键词： Transformer thermostatic oil bath model free adaptive control parameter optimization salp swarm algorithm

来源：评论

学校读者我要写书评

暂无评论

A semi-supervised deep learning fault diagnosis method based on uncertainty estimation and weighted labels 12

A semi-supervised deep learning fault diagnosis method based...

引用

ieee 12th data driven control and learning systems conference (DDCLS)

作者： Huang, Hanxin Zhou, Funa Jia, Pengpeng Wen, Yanqi Shanghai Maritime Univ Sch Logist Engn Shanghai 201306 Peoples R China

ISBN: (纸本)9798350321050

The fault diagnosis method based on deep learning has been widely used due to its powerful feature extraction ability. However, its excellent performance relies on sufficient labeled data participating in model training. Semi-supervised learning is a very effective method to address the issue of low diagnosis accuracy caused by insufficient labeled data. FixMatch, which combines consistent regularization and pseudo-labels, has recently performed well on semi-supervised learning. However, the pseudo-labels of unlabelled samples obtained in FixMatch by high and fixed thresholds can lead to slow model convergence, and such pseudo-labels obtained using pre-defined high thresholds may still be incorrect, and guided by these incorrect pseudo-labels can lead to significant degradation in model performance. Also it is obviously unreasonable to use the same weight for the consistency loss of all unlabelled samples in FixMatch without considering the differences between samples. To address the above issues, a semi-supervised deep learning fault diagnosis method based on uncertainty estimation and weighted labels is designed and experimentally validated using the Case Western Reserve University bearing dataset. Experimental results show that when there are only 2 labels per class, the accuracy of the test set reaches 90.59% and the fault diagnosis accuracy of our proposed method is improved by 9.08% compared to FixMatch.

关键词： fault diagnosis semi-supervised learning uncertainty estimation weighted labels

来源：评论

学校读者我要写书评

暂无评论

Sliding-Mode control Strategy for Dynamic Wireless Charging System with Long Guide Transmitting Coil for EV 12

Sliding-Mode Control Strategy for Dynamic Wireless Charging ...

引用

ieee 12th data driven control and learning systems conference (DDCLS)

作者： Guo, Yan Song, Yang Zhao, Wanqing Shanghai Univ Sch Mech Engn & Automat Shanghai 200444 Peoples R China Shanghai Key Lab Power Stn Automat Technol Shanghai 200444 Peoples R China Newcastle Univ Sch Comp Newcastle Upon Tyne NE4 5TG Tyne & Wear England

ISBN: (纸本)9798350321050

One of the main problems of dynamic wireless charging (DWC) systems for electric vehicles (EVs) proposed in recent years is the variation of mutual inductance caused by the motion of EVs, which leads to the fluctuation of system output voltage. In this paper, the output voltage control problem of EV long guide DWC system is studied. Firstly, the compensation circuit suitable for the system is determined. Secondly, the linear variable parameter model of DWC system is established by considering mutual inductance time varying and combining circuit theory and control theory. Finally, the sliding mode control is applied to the output voltage control of DWC system and the sliding mode controller is designed, so that the output voltage can quickly track the expected reference voltage and improve the anti-position migration ability of the charging system. The simulation results show that the receiver output voltage can track the reference signal well even if the receiver coil has transverse and longitudinal displacement.

关键词： Electric Vehicle Wireless Charging Linear Variable Parameter Model Sliding-Mode controller

来源：评论

学校读者我要写书评

暂无评论

Model Free Adaptive control for Air-Conditioning System in Office Buildings Based on Improved NSGA-II Algorithm 12

Model Free Adaptive Control for Air-Conditioning System in O...

引用

ieee 12th data driven control and learning systems conference (DDCLS)

作者： Chen, Pan Gaol, Jingjing Yang, Xu Zhangl, Tao Univ Sci & Technol Beijing Sch Automat & Elect Engn Beijing 100083 Peoples R China

ISBN: (纸本)9798350321050

driven by the increasing demand for thermal comfort in office building environment and energy saving in building operation process, a model free adaptive control (MFAC) method for air conditioning system based on optimal environmental parameter settings is proposed in this paper. To this end, on the basis of the energy consumption model of common office buildings, the cooling load brought by the fresh air system is considered. The model is combined with the predicted mean vote (PMV) thermal comfort model to constitute an optimization problem. The improved non-dominated sorting genetic algorithm is used to solve the optimization problem to obtain the optimal environmental parameter settings at first. Then, an improved model-free adaptive control method is used for regulation. An adjustment item is added to the estimation algorithm to make the pseudo partial derivative reach the ideal value faster, to reach the optimal environmental parameter setting value faster. Finally, the experimental verification is carried out on the building environment monitoring system of an office building demonstration room, which shows the proposed method can achieve the optimal environment parameter settings, and at the same time can reduce the number of regulation of the air conditioning system and extend the service life of the equipment.

关键词： Office buildings predicted mean vote non-dominated sorting genetic algorithm model-free adaptive 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：