检索结果-内蒙古大学图书馆

63rd IEEE conference on Decision and control, CDC 2024

作者： Bonassi, Fabio Andersson, Carl Mattsson, Per Schon, Thomas B. Uppsala University Department of Information Technology Lägerhyddsvägen 1 Uppsala751 05 Sweden

ISBN: (纸本)9798350316339

In this paper, we discuss the problem of learning state observers for Recurrent Neural network (RNN) black-box models of dynamical systems. State observers are indeed key to designing state-feedback control laws, such as nonlinear Model Predictive control, with satisfactory closed-loop performance. Besides, they can also improve the training procedure of RNN models themselves. Then, we summarize recent developments aimed at jointly learning RNN models and neural network-based state observers, and we propose a new structure based on the recent S 5 architecture. We finally test various observer structures on a p H neutralization process benchmark system, showing the advantages and shortcomings of each architecture. © 2024 IEEE.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

Dynamic Bidirectional-controlled Inverter-Based Grid Optimized by Neural networks

Dynamic Bidirectional-Controlled Inverter-Based Grid Optimiz...

引用

2024 Vehicle Power and Propulsion conference

作者： Alzayed, Mohamad Lemaire, Michel Chaoui, Hicham Massicotte, Daniel Carleton Univ DoE Ottawa ON Canada UQTR DoEC Trois Rivieres PQ Canada Univ Quebec Trois Rivieres Dept Elect & Comp Engn Trois Rivieres PQ Canada Carleton Univ Dept Elect Ottawa ON Canada

ISBN: (纸本)9798331541613;9798331541606

Grid-based voltage source inverters frequently utilize the droop control technique combined with inner/outer voltage and current regulation mechanisms to ensure a dependable electricity supply. This study seeks to introduce a Cascade-Forward Neural network (CFNN) control approach designed to lead inverter-based grids when operating in grid-connected or islanded modes, focusing on improving the transient state performance of the CFNN technique. The suggested approach involves utilizing the inverter in a bidirectional manner, suitable for a diverse array of battery energy storage systems and distributed generation setups. The proposed strategy leverages CFNN to grasp the inverter's non-linear model, enabling precise monitoring of demand and reference power across various operational scenarios within smart grid applications. Furthermore, the approach redefines the grid control concept, guiding the inverter according to optimal parameters that encompass power demand, reference power, equipment dimensions, and external disturbances. Notably, this method circumvents the need for any manual tuning procedures. Furthermore, incorporating dynamic elements into the approach improves the protection system's responsiveness. This ensures continuous power supply during faults, enabling a more effective and rapid response, enhancing system resilience, and reducing downtime. This dual advantage of better power supply and heightened protection system sensitivity highlights the proposed method's significance in fortifying the reliability of power systems.. To assess the efficacy of the suggested CFNN controller, its power tracking, operational capabilities, and dynamic response are assessed via multiple experimental trials employing the hardware-in-the-loop (HIL) approach across various scenarios. The outcomes of these tests are meticulously compared to a well-established conventional strategy, affirming the effectiveness of the proposed method.

关键词： dynamic control inverter-based grid neural network

来源：评论

学校读者我要写书评

暂无评论

Advanced Persistent Threats Detection based on Deep Learning Approach 6

Advanced Persistent Threats Detection based on Deep Learning...

引用

IEEE 6th International conference on Industrial Cyber-Physical systems (ICPS)

作者： Eke, Hope Nkiruka Petrovski, Andrei Robert Gordon Univ Sch Comp Aberdeen Scotland Natl Subsea Ctr Sch Comp Aberdeen Scotland

ISBN: (纸本)9798350311259

Advanced Persistent Threats (APTs) have been a major challenge in securing both Information Technology (IT) and Operational Technology (OT) systems. APT is a sophisticated attack that masquerade their actions to navigates around defenses, breach networks, often, over multiple network hosts and evades detection. It also uses "low-and-slow" approach over a long period of time. Resource availability, integrity, and confidentiality of the operational cyber-physical systems (CPS) state and control is highly impacted by the safety and security measures in place. A framework multi-stage detection approach termed "APT(DASAC)" to detect different tactics, techniques, and procedures (TTPs) used during various APT steps is proposed. Implementation was carried out in three stages: (i) Data input and probing layer - this involves data gathering and pre-processing, (ii) Data analysis layer;applies the core process of "APT(DASAC)" to learn the behaviour of attack steps from the sequence data, correlate and link the related output and, (iii) Decision layer;the ensemble probability approach is utilized to integrate the output and make attack prediction. The framework was validated with three different datasets and three case studies. The proposed approach achieved a significant attacks detection capability of 86.36% with loss as 0.32%, demonstrating that attack detection techniques applied that performed well in one domain may not yield the same good result in another domain. This suggests that robustness and resilience of operational systems state to withstand attack and maintain system performance are regulated by the safety and security measures in place, which is specific to the system in question.

关键词： Advanced Persistent Threats Cyber-Physical systems Critical Infrastructures Deep Learning Industrial control systems Supervisory control and Data Acquisition

来源：评论

学校读者我要写书评

暂无评论

A Multitask and Kernel approach for Learning to Push Objects with a Target-Parameterized Deep Q-network

A Multitask and Kernel approach for Learning to Push Objects...

引用

IEEE/RSJ International conference on Intelligent Robots and systems (IROS)

作者： Ewerton, Marco Villamizar, Michael Jankowski, Julius Calinon, Sylvain Odobez, Jean-Marc Idiap Res Inst Switzerland Martigny Switzerland

ISBN: (纸本)9781665491907

Pushing is an essential motor skill involved in several manipulation tasks, and has been an important research topic in robotics. Recent works have shown that Deep Q-networks (DQNs) can learn pushing policies (when, where to push, and how) to solve manipulation tasks, potentially in synergy with other skills (e.g. grasping). Nevertheless, DQNs often assume a fixed setting and task, which may limit their deployment in practice. Furthermore, they suffer from sparse-gradient backpropagation when the action space is very large, a problem exacerbated by the fact that they are trained to predict state-action values based on a single reward function aggregating several facets of the task, rendering the model training challenging. To address these issues, we propose a multi-head target-parameterized DQN to learn robotic manipulation tasks, in particular pushing policies, and make the following contributions: i) we show that learning to predict different reward and task aspects can be beneficial compared to predicting a single value function where reward factors are not disentangled;ii) we study several alternatives to generalize a policy by encoding the target parameters either into the network layers or visually in the input;iii) we propose a kernelized version of the loss function, allowing to obtain better, faster and more stable training performance. Extensive experiments on simulations validate our design choices, and we show that our architecture learned on simulated data can achieve high performance in a real-robot setup involving a Franka Emika robot arm and unseen objects.

关键词： network layers

来源：评论

学校读者我要写书评

暂无评论

Event-Triggered-Based Robust control of Heterogeneous Connected Vehicles Platoon with Communication Delay and Random Packet Loss 8

Event-Triggered-Based Robust Control of Heterogeneous Connec...

引用

8th CAA International conference on Vehicular control and Intelligence, CVCI 2024

作者： Liu, Chen Ru, Shuangkun Qi, Lin Fang, Jiayi Zhang, Ting Yanshan University Hebei Key Laboratory of Industrial Computer Control Engineering Hebei Province Qinhuangdao066004 China School of Mechanical Engineering Beijing Institute of Technology Beijing100081 China School of Vehicle and Mobility Tsinghua University Beijing100081 China

ISBN: (纸本)9798331504892

With the development of vehicle autonomous driving technology and wireless network communication technology, the mix platoon of connected autonomous vehicles (CAVs) and connected human-driven vehicles (CHVs) is becoming inevitable. The abnormal V2V communication and uncertain vehicle dynamics may affect the CAV control performance. Considering time-varying communication delays and random packet loss, an event-trigger-based robust control strategy is proposed based on the stability theory of stochastic time-delay systems. The Bernoulli packet loss model is constructed to solve the random packet loss problem. The robust controller with the event-triggered mechanism designed using the Lyapunov functional stability theory of stochastic systems ensures the stability of the mix platoon with vehicle heterogeneity and the stability of CAV. The simulation comparison has verified that the proposed control strategy can better suppress the fluctuation of the preceding vehicle and more quickly reach the desired value when both time delay and packet loss are present. © 2024 IEEE.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Advanced Intrusion Detection systems Leveraging Knowledge Graph-Based Techniques 12

Advanced Intrusion Detection Systems Leveraging Knowledge Gr...

引用

2024 12th International conference on systems and control, ICSC 2024

作者： Daoud, Mohamed Amine Mostefaoui, Sid Ahmed Mokhtar Meghazi, Hadj Madani Labbadi, Ourida Zenina, Chaima Bouguessa, Abdelkader Ibn-Khaldoun University of Tiaret LRIAS Lab Dept. of Computer Sciences Algeria Ibn-Khaldoun University of Tiaret Dept. of Computer Sciences Algeria

ISBN: (纸本)9798331531812

The integration of intrusion detection systems (IDS) is crucial for strengthening network security. Improving IDS performance requires advanced techniques for handling intrusion detection data, with machine learning playing a key role. However, machine learning methods often face difficulties in detecting complex attack patterns and are prone to generating a high number of false positives, especially when dealing with unknown attacks. To overcome these issues, enhancing IDS capabilities with knowledge graphs is essential. Knowledge graphs have proven to be valuable tools for modeling and analyzing complex interactions within security data. This article proposes an approach that combines knowledge graphs with machine learning and deep learning. By harnessing the semantic modeling, querying, and reasoning capabilities of knowledge graphs, this integration aims to address the challenges of intelligent detection and decision-making in cybersecurity. © 2024 IEEE.

关键词： network intrusion

来源：评论

学校读者我要写书评

暂无评论

Tariff Versus Sanction Under Bounded Rationality 11th

Tariff Versus Sanction Under Bounded Rationality

引用

11th International conference on network Games, control and Optimization

作者： Pokou, Fredy Fochesato, Marta Le Cadre, Helene Univ Lille Inria CNRS Cent LilleUMR 9189 CRIStAL F-59000 Lille France Swiss Fed Inst Technol Automat Control Lab Dept Elect Engn & Informat Technol Phys Str CH-38092 Zurich Switzerland

ISBN: (纸本)9783031785993;9783031786006

We formulate the design of a taxation mechanism as a Stackelberg game assuming: a) perfect competition, with exogenous prices;b) imperfect competition, captured through a variational inequality approach, with endogenous prices. Three settings of the mechanism are considered: (i) benchmark involving no taxation, (ii) optimum tariff, (iii) optimum sanction. The expected utility maximization formulation of the game is extended further by relying on cumulative prospect theory to account for the bounded rationality of the stakeholders. We derive closed-form mappings linking the outcomes of the three settings. Additionally, we assess the impact of bounded rationality through a new performance metric, the Price of Irrationality. Numerical results are derived on a randomized instance of a gas trading game between Europe, Asia, and Russia.

关键词： Game Theory Sanction Design Cumulative Prospect Theory

来源：评论

学校读者我要写书评

暂无评论

A novel learning approach to MIMO networked control systems with uncertain parameters 14

A novel learning approach to MIMO networked control systems ...

引用

14th IEEE Global Reliability and Prognostics and Health Management conference, PHM-Hangzhou 2023

作者： Jia, Hanguang School of Automation Science and Engineering South China University of Technology National Key Laboratory of Science and Technology on Reliability Physics and Application of Electronic Component The Fifth Electronics Research Institute of the Ministry of Industry and Information Technology Guangzhou China

ISBN: (纸本)9798350301359

This paper proposes a new learning approach to solve control design problems involving uncertain parameters in MEMO networked control systems. This approach combines neural network technology with control system modeling in real-time through learning and adaptation of data, thereby achieving stability and optimization of network-based control systems. The authors first establish a network-based control model and then improve it by replacing the linear parameters with nonlinear ones. To ensure real-time and stable communication between devices, this method also uses adaptive iterative learning control (AILC) technology and error feedback control technology. A network-based control system is used to verify the effectiveness of the proposed approach. The results show that this approach is effective in achieving sufficient communication between devices, adapting to uncertain parameters, and improving control performance. © 2023 IEEE.

关键词： AILC communication error feedback control MIMO networked control systems uncertain parameters

来源：评论

学校读者我要写书评

暂无评论

Can graph neural network-based detection mitigate the impact of hardware imperfections?

Can graph neural network-based detection mitigate the impact...

引用

International Balkan conference on Communications and networking (BalkanCom)

作者： Mitsiou, Lamprini Trevlakis, Stylianos Tsiolas, Argiris Vergados, Dimitrios J. Michalas, Angelos Boulogeorgos, Alexandros-Apostolos A. InnoCube PC Res & Dev Dept Thessaloniki 55535 Greece Univ Western Macedonia Dept Elect & Comp Engn Kozani 50100 Greece

ISBN: (纸本)9798350339109

Until recently, researchers used machine learning methods to compensate for hardware imperfections at the symbol level, indicating that optimum radio-frequency transceiver performance is possible. Nevertheless, such approaches neglect the error correcting codes used in wireless networks, which inspires machine learning (ML)-approaches that learn and minimise hardware imperfections at the bit level. In the present work, we evaluate a graph neural network (GNN)-based intelligent detector's in-phase and quadrature imbalance (IQI) mitigation capabilities. We focus on a high-frequency, high-directional wireless system where IQI affects both the transmitter (TX) and the receiver (RX). The TX uses a GNN-based decoder, whilst the RX uses a linear error correcting algorithm. The bit error rate (BER) is computed using appropriate Monte Carlo simulations to quantify performance. Finally, the outcomes are compared to both traditional systems using conventional detectors and wireless systems using belief propagation based detectors. Due to the utilization of graph neural networks, the proposed algorithm is highly scalable with few training parameters and is able to adapt to various code parameters.

关键词： Belief propagation bit error rate graph neural networks hardware imperfection mitigation in-phase and quadrature imbalance machine learning

来源：评论

学校读者我要写书评

暂无评论

Optimal deployment strategy of airships for near-space communication system using multiobjective advanced evolutionary algorithm

引用

INTERNATIONAL JOURNAL OF COMMUNICATION systems 2024年第3期37卷

作者： Alam, Md. Sarfraz Mondal, Abhishek Natl Inst Technol Silchar Dept Elect & Commun Engn Silchar India

This paper presents a near-space communication system (NSCS) using advanced deployment strategies to gain high throughput. The airships are deployed according to the user's location, assuming robust backbone network characteristics such as signal path loss, fading factor, routing efficiency, and safety issues among bi-connected airships. Due to the independent flying nature of airships, it is very attractive to deploy them as aerial base stations and construct airborne networks to provide service for on-ground users. However, it is quite challenging to optimally deploy multiple airships for on-demand coverage while maintaining the connectivity among airships. A balance between the network parameters, i.e., capacity and coverage area, should be maintained for optimal deployment of the airships. We have derived the maximum throughput of NSCS, including system parameters, as a multiobjective optimization problem subjected to efficient routing protocol and safety constraints. A decomposition-based advanced multiobjective evolutionary algorithm (AMOEA/D) is adopted to solve the deployment optimization problem. The proposed algorithm is motivated by the non-dominated solutions that maintain population diversity over the variable space. Two designed test problems, that is, the L-shaped hotspot problem and nine hotspot problems, are also investigated. Numerical results show that the proposed method improves the system performance compared with benchmark external archive-guided MOEA/D (EGA-MOEA/D) and non-dominated sorted genetic algorithms (NSGA-ii) by 10.46% and 3.84%, respectively. This figure illustrates the system model of the near-space communication system. It consists of multiple airships with the same characteristics following identical communication protocols. All airships hover at the same altitude and serve as aerial base stations at any particular instant. The coverage area of each airship depends on its altitude and communication range. These airships can serve

关键词： advanced multiobjective evolutionary algorithm bi-connected network distributed motion control near-space communication system routing efficiency

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：