检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Xie, Guoyang Wang, Jinbao Yu, Guo Zheng, Feng Jin, Yaochu Department of Computer Science University of Surrey Guildford SurreyGU2 7XH United Kingdom Department of Computer Science Southern University of Science and Technology Shenzhen518055 China Key Laboratory of Advanced Control and Optimization for Chemical Processes Ministry of Education East China University of Science and Technology Shanghai 200237 China Faculty of Technology Bielefeld University Bielefleld D-33615 Germany

Due to limited computational cost and energy consumption, most neural network models deployed in mobile devices are tiny. However, tiny neural networks are commonly very vulnerable to attacks. Current research has proved that larger model size can improve robustness, but little research focuses on how to enhance the robustness of tiny neural networks. Our work focuses on how to improve the robustness of tiny neural networks without seriously deteriorating of clean accuracy under mobile-level resources. To this end, we propose a multiobjective oneshot network architecture search (NAS) algorithm to obtain the best trade-off networks in terms of the adversarial accuracy, the clean accuracy and the model size. Specifically, we design a novel search space based on new tiny blocks and channels to balance model size and adversarial performance. Moreover, since the supernet significantly affects the performance of subnets in our NAS algorithm, we reveal the insights into how the supernet helps to obtain the best subnet under white-box adversarial attacks. Concretely, we explore a new adversarial training paradigm by analyzing the adversarial transferability and the difference between training the subnets from scratch and fine-tuning. Finally, we make a statistical analysis for the layer-wise combination of certain blocks and channels on the first non-dominated front, which can serve as a guideline to design tiny neural network architectures for the resilience of adversarial perturbations. Copyright © 2021, The Authors. All rights reserved.

关键词： Multiobjective optimization

来源：评论

学校读者我要写书评

暂无评论

Youla Parameterization Based control Performance Degradation Online Recovery of Single Inverter control Systems in Microgrids: a PnP Approach

引用

IFAC-PapersOnLine 2020年第2期53卷 12080-12085页

作者： Y. Liu C. Hu H. Lu S.X. Ding L. Li X. Peng Y. Wang Institute for Automatic Control and Complex Systems University of Duisburg-Essen Bismarkstr. 81 BB 47057 Duisburg Germany College of Electrical and Control Engineering North China University of Technology 100144 Beijing China School of Automation and Electrical Engineering University of Science and Technology Beijing China Key Laboratory of Advanced Control and Optimization for Chemical Processes East China University of Science and Technology200237 Shanghai China

In an inverter control system of smart microgrids, the control performance degradation due to load fluctuations, unbalanced loads and other factors, is one of the considerable factors affecting the output power quality. In this paper, a plug-and-play (PnP) technology based on Youla parameterization is proposed to achieve system performance recovery without modifying the existing controller. By ‘plugging-in’ a residual generator and a compensator, a dynamic feed-back of residual signals is embedded into the original inverter control system. In order to realize PnP, an online parameter estimation scheme of the compensator is developed. Finally, an inverter control system is applied to show the effectiveness of the proposed approach.

关键词： Youla parameterization performance degradation online recovery power quality microgrids

来源：评论

学校读者我要写书评

暂无评论

Dynamic Adaptive Event-Triggered Scheme for General Linear Multi-Agent Systems ⁎

引用

IFAC-PapersOnLine 2020年第2期53卷 1789-1794页

作者： Wenying Xu Daniel W.C. Ho Wangli He School of Mathematics Southeast University Nanjing 21189 China Potsdam Institute for Climate Impact Research 14412 Potsdam Germany Department of Mathematics City University of Hong Kong Hong Kong. Key Laboratory of Advanced Control and Optimization for Chemical Processes Ministry of Education East China University of Science and Technology Shanghai 200237 China

This paper proposes a novel dynamic adaptive event-triggered scheme to deal with consensus problems in a class of general linear multi-agent systems. Firstly, a fully distributed event-triggered consensus protocol is proposed by assigning a time-varying coupling weight for each edge. Then by introducing an additional internal dynamic variable, a dynamic adaptive event condition is skillfully constructed and Zeno behavior is also successfully excluded. Based on this, the asymptotic consensus is eventually achieved and the frequency of communication among agents is significantly reduced. Finally, one simulation example is provided to verify the effectiveness of the proposed scheme.

关键词： Consensus event-triggered multi-agent system adaptive control

来源：评论

学校读者我要写书评

暂无评论

When autonomous systems meet accuracy and transferability through AI: A survey

arXiv

引用

arXiv 2020年

作者： Zhang, Chongzhen Wang, Jianrui Yen, Gary G. Zhao, Chaoqiang Sun, Qiyu Tang, Yang Qian, Feng Kurths, Jürgen Key Laboratory of Advanced Control and Optimization for Chemical Process Ministry of Education East China University of Science and Technology Shanghai200237 China School of Electrical and Computer Engineering Oklahoma State University StillwaterOK74075 United States Potsdam Institute for Climate Impact Research Potsdam14473 Germany Institute of Physics Humboldt University of Berlin Berlin12489 Germany

With widespread applications of artificial intelligence (AI), the capabilities of the perception, understanding, decision-making and control for autonomous systems have improved significantly in the past years. When autonomous systems consider the performance of accuracy and transferability, several AI methods, like adversarial learning, reinforcement learning (RL) and meta-learning, show their powerful performance. Here, we review the learning-based approaches in autonomous systems from the perspectives of accuracy and transferability. Accuracy means that a well-trained model shows good results during the testing phase, in which the testing set shares a same task or a data distribution with the training set. Transferability means that when a well-trained model is transferred to other testing domains, the accuracy is still good. Firstly, we introduce some basic concepts of transfer learning and then present some preliminaries of adversarial learning, RL and meta-learning. Secondly, we focus on reviewing the accuracy or transferability or both of them to show the advantages of adversarial learning, like generative adversarial networks (GANs), in typical computer vision tasks in autonomous systems, including image style transfer, image super-resolution, image deblurring/dehazing/rain removal, semantic segmentation, depth estimation, pedestrian detection and person re-identification (re-ID). Then, we further review the performance of RL and meta-learning from the aspects of accuracy or transferability or both of them in autonomous systems, involving pedestrian tracking, robot navigation and robotic manipulation. Finally, we discuss several challenges and future topics for using adversarial learning, RL and meta-learning in autonomous systems. Copyright © 2020, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Data-Driven Model Predictive Monitoring for Dynamic processes

引用

IFAC-PapersOnLine 2020年第2期53卷 98-103页

作者： Qingchao Jiang Huaikuan Yi Xuefeng Yan Xinmin Zhang Jian Huang Key Laboratory of Advanced Control and Optimization for Chemical Processes of Ministry of Education East China University of Science and Technology Shanghai 200237 P.R. China Department of Control Science and Engineering Zhejiang University Hangzhou 310027 P.R. China School of Automation and Electrical Engineering University of Science and Technology Beijing Beijing 100083 P.R. China

process monitoring plays an important role in maintaining favorable process operation conditions and is gaining increasing attention in both academic community and industrial applications. This paper proposes a data-driven model predictive fault detection method to achieve efficient monitoring of dynamic processes. First, a measurement sample is projected into a dominant latent variable subspace that captures main variance of the process data and a residual subspace. Then the dominant latent variable subspace is further decomposed as a dynamic feature subspace and a static feature subspace. A fault detection residual is generated in each subspace, and corresponding monitoring statistic is established. By using the model predictive monitoring scheme, not only the status of a process but also the type of a detected fault, namely a dynamic feature fault or a static feature fault, can be identified. Effectiveness of the proposed data-driven model predictive monitoring scheme is tested on a lab-scale distillation process.

关键词： Model predictive process monitoring data-driven process monitoring dynamic processes canonical correlation analysis fault detection

来源：评论

学校读者我要写书评

暂无评论

Predictive control for Continuous Stirred Tank Reactors

Predictive Control for Continuous Stirred Tank Reactors

引用

2019 International Conference on Informatics, control and Robotics 2019信息学、控制和机器人学国际会议(ICICR2019)

作者： Ioana NASCU Ioan NASCU Wen-li DU Sai GU Key Laboratory of Advanced Control and Optimization for Chemical Processes Ministry of Education East China University of Science and Technology Department of Chemical and Process Engineering University of Surrey Department of Automation Technical University of Cluj-Napoca

The paper focuses on the implementation of a model based predictive control(MBPC) method, for Continuous Stirred Tank Reactors. First, the modelling problem of a single irreversible exothermic reaction, taking place in a perfectly mixed continuously stirred tank reactor(CSTR) is presented. The dynamic model consists of differential material and energy balance equations. The control strategy is investigated and evaluated by performing simulations and analyzing the results. The disturbance rejection capacity of the control system(regulatory control performances) have been tested and compared with those obtained using a classical Proportional Integral Derivative(PID) based controller. The results show that this control strategy has good performances and can be efficiently used to control the CSTR.

关键词： Predictive control PID control chemical reactor process model

来源：评论

学校读者我要写书评

暂无评论

The Role of Surface Termination in Halide Perovskites for Efficient Photocatalytic Synthesis

引用

Angewandte Chemie 2020年第31期132卷 13031-13037页

作者： Dong Y. Li K. Luo W. Zhu C. Guan H. Wang H. Wang L. Deng K. Zhou H. Xie H. Bai Y. Li Y. Chen Q. Experimental Center of Advanced Materials Beijing Key Laboratory of Nanophotonics and Ultrafine Optoelectronic Systems School of Materials Science and Engineering Beijing Institute of Technology Beijing 100081 China School of Chemistry and Chemical Engineering Southwest Petroleum University Chengdu 610500 China Key Laboratory of Agro-products Quality and Safety Control in Storage and Transport Process Ministry of Agriculture and Rural Affairs/Institute of Food Science and Technology Chinese Academy of Agricultural Sciences Beijing 100193 China Department of Materials Science and Engineering College of Engineering Peking University Beijing 100871 China School of Physics and Electronics Central South University Changsha Hunan 410083 China

Halide perovskites have received attention in the field of photocatalysis owing to their excellent optoelectronic properties. However, the semiconductor properties of halide perovskite surfaces and the influence on photocatalytic performance have not been systematically clarified. Now, the conversion of triose (such as 1,3-dihydroxyacetone (DHA)) is employed as a model reaction to explore the surface termination of MAPbI3. By rational design of the surface termination for MAPbI3, the production rate of butyl lactate is substantially improved to 7719 μg g−1 cat. h−1 under visible-light illumination. The MAI-terminated MAPbI3 surface governs the photocatalytic performance. Specially, MAI-terminated surface is susceptible to iodide oxidation, which thus promotes the exposure of PbII as active sites for this photocatalysis process. Moreover, MAI-termination induces a p-doping effect near the surface for MAPbI3, which facilitates carrier transport and thus photosynthesis. © 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

关键词： halide perovskite iodide vacancies photocatalysis surface termination triose

来源：评论

学校读者我要写书评

暂无评论

A data-approach for industrial utility systems optimization under uncertainty

引用

Energy 2019年 182卷 559-569页

作者： Zhao, Liang You, Fengqi Key Laboratory of Advanced Control and Optimization for Chemical Process Ministry of Education East China University of Science and Technology Shanghai200237 China Robert Frederick Smith School of Chemical and Biomolecular Engineering Cornell University IthacaNY14853 United States

Energy optimization of utility system helps to reduce the operating cost and save energy for the industrial plants. Widespread uncertainties such as device efficiency and process demand pose new challenges for this issue. A hybrid modeling framework is presented by introducing the operating data into mechanism model to adapt the changes of device efficiency and operating conditions. Mathematical models of boilers, steam turbines, and letdown valves are then developed in the framework. Based on the process historical data of a real-world plant, a Dirichlet process mixture model is used to capture the support information of uncertain parameters. Bridging data-driven robust optimization (DDRO) and utility system optimization under uncertainty, a robust mixed-integer nonlinear programming (MINLP) model is developed by utilizing the derived uncertainty set. The robust counterpart of the developed model can be reformulated as a tractable MINLP problem including conic quadratic constraints that could be solved efficiently. A real-world case study is carried out to demonstrate the effectiveness of the proposed approach in protecting against uncertainties and achieving a good trade-off between optimality and robustness of the operational decisions for industrial utility systems. © 2019 Elsevier Ltd

关键词： Boilers

来源：评论

学校读者我要写书评

暂无评论

Sliding Mode control for Networked control System Under Fading Channels

Sliding Mode Control for Networked Control System Under Fadi...

引用

control Technology and Applications (CCTA),

作者： Jiarui Li Yugang Niu Yuanyuan Zou Key Laboratory of Advanced Control and optimization for Chemical Process (East China University of Science and Technology) Ministry of Education Shanghai China Department of Automation Shanghai Jiao Tong University Shanghai China

In this work, the discrete-time sliding mode control problem for the networked control systems under fading channels is investigated. Moreover, the state signals will be quantized through a logarithmic quantizer before transmitted to the controller. Subsequently, a sliding surface relating to fading coefficient is designed to cope with the fading measurements. It is shown that the designed sliding mode controller can drive the state trajectories into the neighborhood of the specified sliding surface and the closed-loop system is asymptotically stable in the sense of mean square. In the end, a numerical simulation is given.

关键词： Fading channels Quantization (signal) Symmetric matrices Sliding mode control Closed loop systems Lyapunov methods Networked control systems

来源：评论

学校读者我要写书评

暂无评论

Impulsive consensus of leader-following nonlinear multi-agent systems under DoS attacks

Impulsive consensus of leader-following nonlinear multi-agen...

引用

Annual Conference of the IEEE Industrial Electronics Society

作者： Yiting Wang Wangli He Ministry of Education The Key Laboratory of Advanced Control and Optimization for Chemical Processes East China University of Science and Technology Shanghai China

This paper investigates impulsive consensus of nonlinear multi-agent systems in the leader-following framework subject to denial-of-service (DoS) attacks. The case that the communication network suffers from DoS attacks is considered, which will destroy the communication link, resulting in switching topologies. Based on the assumption that the system can recover from DoS attacks, sufficient conditions on the design of the impulse interval and the impulse attack ratio that the system can sustain is provided. A numerical example is given to validate the effectiveness of our theoretical results.

关键词： Multi-agent systems Topology Switches Denial-of-service attack Network topology Communication networks

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：