检索结果-内蒙古大学图书馆

A Reinforcement learning Method for Constraint-Satisfied Services Composition

IEEE TRANSACTIONS ON SERVICES COMPUTING 2020年第5期13卷 786-800页

作者： Ren, Lifang Wang, Wenjian Xu, Hang Shanxi Univ Sch Comp & Informat Technol Taiyuan 030006 Peoples R China Shanxi Univ Finance & Econ Sch Appl Math Taiyuan 030006 Peoples R China

With increasing adoption and presence of Web services, service composition becomes an effective way to construct software applications. Composite services need to satisfy both the functional and the non-functional requirements. Traditional methods usually assume that the quality of service (qoS) and the behaviors of services are deterministic, and they execute the composite service after all the component services are selected. It is difficult to guarantee the satisfaction of user constraints and the successful execution of the composite service. This paper models the constraint-satisfied service composition (CSSC) problem as a Markov decision process (MDP), namely CSSC-MDP, and designs a q-learning algorithm to solve the model. CSSC-MDP takes the uncertainty of qoS and service behavior into account, and selects a component service after the execution of previous services. Thus, CSSC-MDP can select the globally optimal service based on the constraints which need the following services to satisfy. In the case of selected service failure, CSSC-MDP can timely provide the optimal alternative service. Simulation experiments show that the proposed method can successfully solve the CSSC problem of different sizes. Comparing with three representative methods, CSSC-MDP has obvious advantages, especially in terms of the success rate of service composition.

关键词： quality of service Internet Time factors Optimization Markov processes Uncertainty learning (artificial intelligence) Web service composition constraint-satisfied uncertainty of service behaviors undetermined qoS Markov decision process (MDP) q-learning algorithm

来源：评论

学校读者我要写书评

暂无评论

PID Controller Autotuning Design by a Deterministic q-SLP algorithm

引用

IEEE ACCESS 2020年 8卷 50010-50021页

作者： Pongfai, Jirapun Su, Xiaojie Zhang, Huiyan Assawinchaichote, Wudhichai King Mongkuts Univ Technol Thonburi Dept Elect & Telecommun Engn Fac Engn Bangkok 10140 Thailand Chongqing Univ Coll Automat Chongqing 400044 Peoples R China Chongqing Technol & Business Univ Natl Res Base Intelligent Mfg Serv Chongqing 400067 Peoples R China

The proportional integral and derivative (PID) controller is extensively applied in many applications. However, three parameters must be properly adjusted to ensure effective performance of the control system: the proportional gain and derivative gain (inline-formula). Therefore, the aim of this paper is to optimize and improve the stability, convergence and performance in autotuning the PID parameter by using a deterministic q-SLP algorithm. The proposed method is a combination of the swarm learning process (SLP) algorithm and q-learning algorithm. The q-learning algorithm is applied to optimize the weight updating of the SLP algorithm based on the new deterministic rule and closed-loop stabilization of the learning rate. To validate the global optimization of the deterministic rule, it is proven based on the Bellman equation, and the stability of the learning process is proven with respect to the Lyapunov stability theorem. Additionally, to demonstrate the superiority of the performance and convergence in autotuning the PID parameter, simulation results of the proposed method are compared with those based on the central position control (CPC) system using the traditional SLP algorithm, the whale optimization algorithm (WOA) and improved particle swarm optimization (IPSO). The comparison shows that the proposed method can provide results superior to those of the other algorithms with respect to both performance indices and convergence.

关键词： Convergence Optimization Stability analysis Prediction algorithms Mathematical model Simulation Tuning Autotuning gain central position control system q-learning algorithm PID controller swarm learning process algorithm optimal control

来源：评论

学校读者我要写书评

暂无评论

TABU SEARCH GUIDED BY REINFORCEMENT learning FOR THE MAX-MEAN DISPERSION PROBLEM

引用

JOURNAL OF INDUSTRIAL AND MANAGEMENT OPTIMIZATION 2021年第6期17卷 3223-3246页

作者： Nijimbere, Dieudonne Zhao, Songzheng Gu, Xunhao Esangbedo, Moses Olabhele Dominique, Nyiribakwe Northwestern Polytech Univ Sch Management Xian 710072 Peoples R China Xidian Univ Dept Comp Sci & Technol Xian 710071 Peoples R China

We present an effective hybrid metaheuristic of integrating reinforcement learning with a tabu-search (RLTS) algorithm for solving the max- mean dispersion problem. The innovative element is to design using a knowledge strategy from the q-learning mechanism to locate promising regions when the tabu search is stuck in a local optimum. Computational experiments on extensive benchmarks show that the RLTS performs much better than stateof-the-art algorithms in the literature. From a total of 100 benchmark instances, in 60 of them, which ranged from 500 to 1,000, our proposed algorithm matched the currently best lower bounds for all instances. For the remaining 40 instances, the algorithm matched or outperformed. Furthermore, additional support was applied to present the effectiveness of the combined RL technique. The analysis sheds light on the effectiveness of the proposed RLTS algorithm.

关键词： Hybrid metaheuristics algorithm reinforcement learning q-learning algorithm tabu search

来源：评论

学校读者我要写书评

暂无评论

learning-based secure communication against active eavesdropper in dynamic environment

引用

IET COMMUNICATIONS 2019年第15期13卷 2235-2242页

作者： He, Dongxuan Wang, Hua Zhou, He Beijing Inst Technol Sch Informat & Elect Beijing Peoples R China

In this study, the authors propose a learning-based approach to improve the security of the authors' considered communication system in a dynamic environment, where a source transmits information to a legitimate receiver in the presence of an active eavesdropper. Additionally, they assume that the source has to harvest energy from the environment to support its communication. Due to the dynamic of the environment, both the harvested energy and the channel vary over time, requiring a dynamic transmission strategy that follows the changes. In order to improve the security performance, they first analyse how to select the optimal transmission parameters in hindsight, and then they propose to combine the q-learning algorithm and the expert advice method to maximise the cumulative reward in the dynamic environment. They also introduce an improved learning-based approach, which accelerates the convergence of their approach. The simulation results show that their proposed learning-based approach helps the legitimate nodes learn a beneficial transmission strategy to obtain a larger cumulative reward.

关键词： telecommunication security radio networks learning (artificial intelligence) authors active eavesdropper secure communication improved learning-based approach dynamic environment q-learning algorithm security performance dynamic transmission strategy harvested energy

来源：评论

学校读者我要写书评

暂无评论

Embedding a priori knowledge in reinforcement learning

引用

JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS 1998年第1期21卷 51-71页

作者： Ribeiro, CHC Univ London Imperial Coll Sci Technol & Med Dept Elect & Elect Engn London SW7 2BT England

In the last years, temporal differences methods have been put forward as convenient tools for reinforcement learning. Techniques based on temporal differences, however, suffer from a serious drawback: as stochastic adaptive algorithms, they may need extensive exploration of the state-action space before convergence is achieved. Although the basic methods are now reasonably well understood, it is precisely the structural simplicity of the reinforcement learning principle learning through experimentation - that causes these excessive demands on the learning agent. Additionally, one must consider that the agent is very rarely a tabula rasa: some rough knowledge about characteristics of the surrounding environment is often available. In this paper, I present methods for embedding a priori knowledge in a reinforcement learning technique in such a way that both the mathematical structure of the basic learning algorithm and the capacity to generalise experience across the state-action space are kept. Extensive experimental results show that the resulting variants may lead to good performance, provided a sensible balance between risky use of prior imprecise knowledge and cautious use of learning experience is adopted.

关键词： q-learning algorithm reinforcement learning experience generalisation

来源：评论

学校读者我要写书评

暂无评论

A smart home management system with hierarchical behavior suggestion and recovery mechanism

引用

COMPUTER STANDARDS & INTERFACES 2015年 41卷 98-111页

作者： Shen, Victor R. L. Yang, Cheng-Ying Chen, Chien Hung Natl Taipei Univ Coll Elect Engn & Comp Sci Dept Comp Sci & Informat Engn New Taipei City 237 Taiwan Univ Taipei Dept Comp Sci Taipei 100 Taiwan Natl Taipei Univ Grad Inst Elect Engn Coll Elect Engn & Comp Sci New Taipei City 237 Taiwan

We propose the hierarchical behavior suggestion system and recovery mechanism for the smart home management platform, including location layer, action layer, and home appliance layer. The smart home management system uses the hierarchical structure to take regional management action and home appliance management action. This study also provides a hierarchical human behavior suggestion algorithm (HHBSA), which suggests the behavior pattern. HHBSA includes a location-learning suggestion algorithm (LISA) and an action-behavior suggestion algorithm (ABSA). LISA suggests the user's location with the concepts of q-learning and fuzzy-state q-learning (FSqL). ABSA provides advices on regional behaviors according to the suggested regional sequence updated by users' location. The home appliances included in the behaviors can be switched on in advance when the behaviors have been suggested. A hierarchical recovery mechanism may be used to correct the errors occurring when starting the home appliances. The home appliances can be re-started when errors occur if the action layer is set as a recovery point that can be changed according to the usage sequence. A dynamic recovery point makes it possible to unlimitedly add behaviors to the system, and to maintain the efficiency of a recovery mechanism. (C) 2015 Elsevier B.V. All rights reserved.

关键词： Smart home q-learning algorithm Hierarchical structure Recovery mechanism

来源：评论

学校读者我要写书评

暂无评论

Tool Path Optimization for Complex Cavity Milling Based on Reinforcement learning Approach

引用

IEEE ACCESS 2023年 11卷 66793-66807页

作者： Wan, Yi Xu, Wei Zuo, Tian-Yu Nanjing Xiaozhuang Univ Sch Environm Sci Nanjing 211171 Peoples R China Sanjiang Univ Sch Mech & Elect Engn Nanjing 210012 Peoples R China Nanjing Univ Informat Sci & Technol Sch Automat Nanjing 210044 Peoples R China

In the machining of parts, tool paths for complex cavity milling often have different generation options, as opposed to simple machining features. The different tool path generation options influence the machining time and cost of the part during the machining process. Decision makers prefer tool path solutions that have fewer blanking lengths, which means that the machining process is more efficient. Therefore, in order to reduce costs and increase efficiency, it is necessary to carefully design the tool path generation for the features to be machined on the part, especially for complex cavity milling features. However, solutions to the problem of optimal design of tool paths for complex cavity milling features have not been well developed in current research work. In this paper, we present a systematic solution for complex cavity milling tool path generation based on reinforcement learning. First, a grid converter is executed for converting the 3D geometry of the cavity milling feature into a matrix of planar grid points recognisable by the program, set according to the cutting parameters. Afterwards, the tool path generation process is refined and modelled as a Markov decision process. Ultimately, a tool path generation solution combining the A* algorithm with the q-learning algorithm is executed. The agent iterates through trial and error to construct an optimal tool path for a given cavity milling task. Three case experiments demonstrate the feasibility of the proposed approach. The superiority of the reinforcement learning-based approach in terms of solution speed and solution quality is further demonstrated by comparing the proposed approach with the evolutionary computational techniques currently popular in research for solving tool path optimisation design problems.

关键词： Tool path optimization cavity milling path planning algorithm reinforcement learning q-learning algorithm

来源：评论

学校读者我要写书评

暂无评论

learning-Based Modeling and Optimization for Real-Time System Availability

引用

IEEE TRANSACTIONS ON COMPUTERS 2021年第4期70卷 581-594页

作者： Li, Liying Zhou, Junlong Wei, Tongquan Chen, Mingsong Hu, Xiaobo Sharon East China Normal Univ Sch Comp Sci & Technol Engn Res Ctr Software Hardware Codesign Technol & Shanghai 200062 Peoples R China Nanjing Univ Sci & Technol Sch Comp Sci & Engn Nanjing 210094 Peoples R China Univ Notre Dame Dept Comp Sci & Engn Notre Dame IN 46656 USA

As the density of integrated circuits continues to increase, the possibility that real-time systems suffer from soft and hard errors rises significantly, resulting in a degraded availability of system. In this article, we investigate the dynamic modeling of cross-layer soft error rate based on the Back Propagation (BP) neural network, and propose optimization strategies for system availability based on Cross Entropy (CE) and q-learning algorithms. Specifically, the BP neural network is trained using cross-layer simulation data obtained from SPICE simulation while the optimization for system availability is achieved by judiciously selecting an optimal supply voltage for processors under timing constraints. Simulation results show that the CE-based method can improve system availability by up to 32 percent compared to state-of-the-art methods, and the q-learning-based algorithm can further enhance system availability by up to 20 percent compared to the proposed CE-based method.

关键词： Error analysis Integrated circuit modeling Task analysis Optimization Neural networks Real-time systems Data models BP neural network cross entropy cross-layer modeling q-learning algorithm system availability soft and hard errors

来源：评论

学校读者我要写书评

暂无评论

A reinforcement learning approach using Markov decision processes for battery energy storage control within a smart contract framework

引用

JOURNAL OF ENERGY STORAGE 2024年 86卷

作者： Jonban, Mansour Selseleh Romeral, Luis Marzband, Mousa Abusorrah, Abdullah Univ Politecn Cataluna Elect Engn Dept MCIA Ctr Terrassa Spain King Abdulaziz Univ Ctr Res Excellence Renewable Energy & Power Syst Jeddah 21589 Saudi Arabia King Abdulaziz Univ Fac Engn KA Care Energy Res & Innovat Ctr Dept Elect & Comp EngnRenewable Energy & Power Sy Jeddah 21589 Saudi Arabia

With the increasing penetration of renewable energy sources (RESs), the necessity for employing smart methods to control and manage energy has become undeniable. This study introduces a real -time energy management system based on a multi-agent system supervised by a smart contract, employing a bottom -up approach for a grid-connected DC micro-grid equipped with solar photovoltaic panels (PV), wind turbines (WT), microturbines (MT), and battery energy storage (BES). Each unit is controlled and managed through a distributed decision-making structure. The BES agent is governed by an intelligent structure based on a reinforcement learning model. To facilitate interaction and coordination among agents, a tendering process is employed, where each agent, under its supervised control structure, presents its offer for the tendered item at each time period. The tendering organization allocates demand using the first-price sealed-bid algorithm among bidders to optimize energy cost in the Microgrid. The proposed approach offers a real -time intelligent system capable of ensuring fault tolerance, stability, and reliability in the Microgrid. The main achievement of this study is the development of a robust real -time energy management system that integrates various renewable energy sources and battery storage while ensuring efficient operation and resilience in the face of faults or disruptions.

关键词： Energy management system Multi-agent system Smart contract Tendering process First-price sealed-bid algorithm Reinforcement learning Markov decision process q-learning algorithm Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

Simulation Model for the AGC System of Isolated Microgrid Based on q-learning Method

Simulation Model for the AGC System of Isolated Microgrid Ba...

引用

IEEE Data Driven Control and learning Systems Conference

作者： Penghu Wang Hao Tang Kai Lv School of Electrical Engineering and Automation Hefei University of Technology

ISBN: (纸本)9781538626191

The automatic generation control (AGC) in isolated microgrid with multiple distributed energy resources is concerned in this study. First, the load frequency control (LFC) model of an isolated microgrid, which contains diesel engine generators, super-magnetic magnetic energy storage, wind turbines and photovoltaic power system, is established through the analysis of the power generation characteristics of each distributed generation (DG). The LFC model of an isolated microgrid is built by MATLAB/Simulink with diesel generators as frequency control units. Based on the AGC principle of power grid, the AGC controller of the microgrid system is designed by the q learning algorithm based on the discount compensation model to complete the frequency control. The simulation results verify the feasibility of the isolated microgrid model, showing the efficient dynamic performance of q controller by compared with PI controller.

关键词： Microgrid Distributed generation Automatic generation control Load frequency control q-learning algorithm

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：