检索结果-内蒙古大学图书馆

5th International Workshop on Advanced Computational Intelligence and Intelligent Informatics, IWACIII 2017

作者： Zhang, Jialiang An, Jianqi Wu, Min Cao, Weihua School of Automation China University of Geosciences Wuhan430074 China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan430074 China Department of Computer Science School of Computing Tokyo Institute of Technology Yokohama226-8502 Japan

The development and distribution states of the blast furnace (BF) gas flow determine whether the reaction in the furnace can proceed normally and decide the utilization efficiency of the carbon in the furnace and the stable smooth operation of the furnace. The internal environment of BF is closed with high temperature and high pressure, which causes that the relationship between the development situation of edge gas flow of the BF and the BF operation is uncertain. In this paper, a prediction model of edge gas flow development state based on BP neural network (BPNN) is proposed, clarifying the relationship between the development state of edge gas flow and the edge burden parameters. Firstly, this paper gets the relative operation factors of effecting the development situation of edge gas flow through analyzing the process mechanism of the BF ironmaking, and proposes the development index of edge gas flow and the heat load of cooling stave to measure the development state of the edge gas flow. Secondly, the intrinsic correlation between the data is verified, and the data are analyzed by Spearman's rank correlation coefficient (SRCC) analysis. Finally, according to the analysis results, combined with the burden mechanism, the edge gas flow prediction model is established to realize the prediction of the development state of the edge gas flow. The calculation of the data collected in the metallurgy scene of the BF suggested that this prediction model of the edge gas flow development state of the BF could describe the relationship between BF operation parameters and the developing state of the edge gas flow.

关键词： Flow of gases

来源：评论

学校读者我要写书评

暂无评论

Cooperative differential evolution with dynamical population for short-term traffic flow prediction problem

引用

International Journal of Performability Engineering 2018年第4期14卷 785-794页

作者： Wang, Danping Hu, Kunyuan He, Maowei Chen, Hanning Department of Information Service and Intelligent Control Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China University of Chinese Academy of Sciences Beijing100039 China Shenyang University Shenyang110044 China School of Computer Science and Software Tianjin Polytechnic University Tianjin300387 China

Differential Evolution (DE) is a heuristic stochastic search algorithm based on population differences, which has advantages of simple parameters and fast convergence rate. However, it has weak robustness, especially for multimodal problems. Therefore, this paper proposes a Cooperative Differential Evolution with Dynamical population (DynCDE). In the proposed algorithm, the K-means method is employed to partition the whole population. For the high convergence rate of DE/current-to-best/ 1/bin, the neighbor-based mutation strategy is applied and the dynamic population size method based on aging mechanism and lifecycle mechanism is designed to keep the balance between exploration and exploitation. This modified DE has the potential to improve prediction accuracy of neural networks. Finally, this DynCDE-based neural network model is applied to solving the short-term traffic flow prediction problem, which offers very excellent results. © 2018 Totem Publisher, Inc. All rights reserved.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

PID2018 benchmark challenge: Model-based feedforward compensator with a conditional integrator ??

arXiv

引用

arXiv 2018年

作者： Yuan, Jie Ates, Abdullah Dehghan, Sina Zhao, Yang Fei, Shumin Chen, YangQuan School of Automation Southeast University Nanjing210096 China Engineering Faculty Computer Engineering Department Inonu University Malatya44280 Turkey Mechatronics Embedded Systems and Automation Lab University of California MercedCA95340 United States School of Science and Engineering Shandong University Jinan250061 China Key Laboratory of Measurement and Control of CSE Ministry of Education School of Automation Southeast University Nanjing210096 China

Since proportional-integral-derivative (PID) controllers absolutely dominate the control engineering, numbers of different control structures and theories have been developed to enhance the effciency of PID controllers. Thus, it is essential and inspiring to operate different PID control strategies to the PID2018 Benchmark Challenge. In this paper, a novel control strategy is designed for this refrigeration system, where a feedforward compensator and a conditional integrator are utilized to compensate the disturbances and remove the steady-state error in the benchmark problem, respectively. The simulation results given in the benchmark problem show the straightforward effectiveness of the proposed control structure compared with the existing control methods. Copyright © 2018, The Authors. All rights reserved.

关键词： Disturbance rejection

来源：评论

学校读者我要写书评

暂无评论

The Robust Exact Differentiator Toolbox: Improved Discrete-Time Realization

The Robust Exact Differentiator Toolbox: Improved Discrete-T...

引用

International Workshop on Variable Structure Systems, VSS

作者： M. Reichhartinger S. Koch H. Niederwieser S.K. Spurgeon Institute of Automation and Control Graz University of Technology Graz Austria Christian Doppler Laboratory for Model Based Control of Complex Test Bed Systems Graz University of Technology Graz Austria Information and Computer Engineering Graz University of Technology Graz Austria Department of Electronic & Electrical Engineering University College London London UK

This paper presents a new release of A Robust Exact Differentiator Toolbox for Matlab ® /Simulink ® proposed in [1]. This release features a new discrete-time realization of the continuous-time robust exact differentiator. The implemented discretization scheme is less sensitive to gain overestimation and does not suffer from the discretization chattering effect. Hence, the single tuning parameter of the new version of the implemented differentiator is more intuitive to tune. Furthermore, it shows superior estimation performance in the case of large sampling times in comparison to the previous release. This is confirmed by the presented results obtained by numerical simulations and a real world application.

关键词： Eigenvalues and eigenfunctions Tuning Estimation error Robustness Mathematical model Real-time systems

来源：评论

学校读者我要写书评

暂无评论

Global control for a class of uncertain upper-triangular nonlinear systems with uncontrollable linearization 36

Global control for a class of uncertain upper-triangular non...

引用

第36届中国控制会议

作者： Zha Wenting Qian Chunjiang Zhai Junyong Fei Shumin School of Mechanical Electronic & Information Engineering China University of Mining & Technology (Beijing) Key Laboratory of Measurement and Control of CSE Ministry of EducationSchool of AutomationSoutheast University Department of Electrical and Computer Engineering The University of Texas at San Antonio

ISBN: (纸本)9781538629185

This paper fucuses on the global control problem for a class of upper-triangular nonlinear systems whose linearization around the origin is not guaranteed to be controllable. Assuming that the nonlinearities satisfy the homogeneous growth conditions, a nonsmooth state-feedback controller is elaborately constructed based on the adding a power integrator technique and the homogeneous domination approach. By introducing the dynamic gain to deal with the input dependent growth rate, it can be proved that all the signals of the closed-loop system are bounded and the system states converge to the origin asymptotically.

关键词： upper-triangular nonlinear systems input dependent growth rate adding a power integrator homogeneous domination approach

来源：评论

学校读者我要写书评

暂无评论

Stability and Performance Analysis of Saturated Systems Using an Enhanced Max Quadratic Lyapunov Function

引用

IFAC-PapersOnLine 2017年第1期50卷 11847-11852页

作者： Li Y. Lin Z. Li N. Department of Automation Shanghai Jiao Tong University Key Laboratory of System Control and Information Processing of Ministry of Education Shanghai 200240 China Charles L. Brown Department of Electrical and Computer Engineering University of Virginia Charlottesville VA 22904-4743 United States

This paper revisits the problem of estimating the domain of attraction and the nonlinear L2 gain of a saturated system with an algebraic loop. A max quadratic Lyapunov function, the maximum function of a group of quadratic Lyapunov functions, is proposed for use in the stability and performance analysis for a saturated system. The max quadratic Lyapunov function involves a partitioning of the state space. Exploiting the special properties of regions of the state space, we propose in this paper an enhanced max quadratic Lyapunov function, which results from adding a term that characterizes the regions of the state space to the max quadratic Lyapunov function. The matrix associated with the enhanced max quadratic Lyapunov function is not required to be positive definite, and thus a set of less conservative stability and performance conditions are established, from which a larger estimate of the domain of attraction and a tighter estimate of the nonlinear L2 gain can be obtained. Simulation results indicate the effectiveness and superiority of the proposed method. © 2017

关键词： Domain of attraction enhanced max quadratic Lyapunov functions input saturation L2 gain

来源：评论

学校读者我要写书评

暂无评论

Two-Layer Distributed Formation-Containment control of Multiple Euler-Lagrange Systems by Output Feedback

引用

IEEE Transactions on Cybernetics 2018年第2期49卷 675-687页

作者： Li, Dongyu Zhang, Wei He, Wei Li, Chuanjiang Ge, Shuzhi Sam Department of Control Science and Engineering Harbin Institute of Technology Harbin150001 China Shanghai Institute of Satellite Engineering Shanghai Key Laboratory of Deep Space Exploration Technology Shanghai201109 China School of Automation and Electrical Engineering University of Science and Technology Beijing Beijing100083 China Department of Electrical and Computer Engineering National University of Singapore Singapore117576 Singapore

This paper addresses the distributed formation-containment (DFC) problem for multiple Euler-Lagrange systems with model uncertainties via output feedback in both constant and time-varying formation cases. First, a novel definition of the DFC problem is proposed using a two-layer framework. Since only parts of the followers can acquire the states of the dynamic leader, we design a distributed finite-time sliding-mode estimator to obtain accurate estimations of the desired position and velocity for each agent. Next, to deal with the absence of velocity sensors, we propose two DFC control laws combined with the high-gain observer for the leaders and the followers, respectively, while the time-varying formation in the first layer and the leader-based containment in the second layer can be achieved. Further, the adaptive neural networks are applied to deal with the model uncertainties due to their superior approximation capability. The uniform ultimate boundedness of all the state errors can be guaranteed by Lyapunov stability theory. In addition, a unified framework is given which can be transformed to four other basic distributed problems. Finally, simulation examples are presented to illustrate the feasibility of the theoretical results. © 2013 IEEE.

关键词： Lagrange multipliers

来源：评论

学校读者我要写书评

暂无评论

Distributed Least Squares Solver for Network Linear Equations

arXiv

引用

arXiv 2018年

作者： Yang, Tao George, Jemin Qin, Jiahu Yi, Xinlei Wu, Junfeng State Key Laboratory of Synthetical Automation for Process Industries Northeastern University Shenyang110819 China U.S. Army Research Laboratory AdelphiMD20783 United States Department of Automation University of Science and Technology of China Hefei230027 China Division of Decision and Control Systems School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm100 44 Sweden College of Control Science and Engineering Zhejiang University Hangzhou310027 China

In this paper, we study the problem of finding the least square solutions of over-determined linear algebraic equations over networks in a distributed manner. Each node has access to one of the linear equations and holds a dynamic state. We first propose a distributed least square solver over connected undirected interaction graphs and establish a necessary and sufficient on the step-size under which the algorithm exponentially converges to the least square solution. Next, we develop a distributed least square solver over strongly connected directed graphs and show that the proposed algorithm exponentially converges to the least square solution provided the step-size is sufficiently small. Moreover, we develop a finite-time least square solver by equipping the proposed algorithms with a finite-time decentralized computation mechanism. The theoretical findings are validated and illustrated by numerical simulation examples. Copyright © 2018, The Authors. All rights reserved.

关键词： Linear equations

来源：评论

学校读者我要写书评

暂无评论

The structure transfer machine theory and applications

arXiv

引用

arXiv 2018年

作者： Yang, Wankou Zhang, Baochang Wang, Ze Zhuo, Lian Han, Jungong Zhen, Xiantong School of Automation Southeast University Key Laboratory of Measurement and Control of CSE Ministry of Education Nanjing210096 China Beihang University Beijing China Shenzhen Academy of Aerospace Technology Shenzhen100083 China Key Laboratory of Measurement and Control of CSE Ministry of Education Nanjing210096 China Department of Computer Science and Digital Technologies Northumbria University Newcastle United Kingdom

Representation learning is a fundamental but challenging problem, especially when the distribution of data is unknown. In this paper, we propose a new representation learning method, named Structure Transfer Machine (STM), which enables feature learning process to converge at the representation expectation in a probabilistic way. We theoretically show that such an expected value of the representation (mean) is achievable if the manifold structure can be transferred from the data space to the feature space. The resulting structure regularization term, named manifold loss, is incorporated into the loss function of the typical deep learning pipeline. The STM architecture is constructed to enforce the learned deep representation to satisfy the intrinsic manifold structure from the data, which results in robust features that suit various application scenarios, such as digit recognition, image classification and object tracking. Compared with state-of-the-art CNN architectures, we achieve better results on several commonly used piblic benchmarks. Copyright © 2018, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Learning ReLU networks on linearly separable data: Algorithm, optimality, and generalization

arXiv

引用

arXiv 2018年

作者： Wang, Gang Giannakis, Georgios B. Chen, Jie Digital Technology Center Department of Electrical and Computer Engineering University of Minnesota MinneapolisMN55455 United States State Key Lab of Intelligent Control Decision of Complex Systems School of Automation Beijing Institute of Technology Beijing100081 China Tongji University Shanghai200092 China

Neural networks with REctified Linear Unit (ReLU) activation functions (a.k.a. ReLU networks) have achieved great empirical success in various domains. Nonetheless, existing results for learning ReLU networks either pose assumptions on the underlying data distribution being e.g. Gaussian, or require the network size and/or training size to be sufficiently large. In this context, the problem of learning a two-layer ReLU network is approached in a binary classification setting, where the data are linearly separable and a hinge loss criterion is adopted. Leveraging the power of random noise perturbation, this paper presents a novel stochastic gradient descent (SGD) algorithm, which can provably train any single-hidden-layer ReLU network to attain global optimality, despite the presence of infinitely many bad local minima, maxima, and saddle points in general. This result is the first of its kind, requiring no assumptions on the data distribution, training/network size, or initialization. Convergence of the resultant iterative algorithm to a global minimum is analyzed by establishing both an upper bound and a lower bound on the number of non-zero updates to be performed. Moreover, generalization guarantees are developed for ReLU networks trained with the novel SGD leveraging classic compression bounds. These guarantees highlight a key difference (at least in the worst case) between reliably learning a ReLU network as well as a leaky ReLU network in terms of sample complexity. Numerical tests using both synthetic data and real images validate the effectiveness of the algorithm and the practical merits of the theory. Copyright © 2018, The Authors. All rights reserved.

关键词： Stochastic systems

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：