检索结果-内蒙古大学图书馆

Reinforcement learning-Based Energy Management Strategy for a Hybrid Electric Tracked Vehicle

ENERGIES 2015年第7期8卷 7243-7260页

作者： Liu, Teng Zou, Yuan Liu, Dexing Sun, Fengchun Beijing Inst Technol Sch Mech Engn Collaborat Innovat Ctr Elect Vehicles Beijing Beijing 100081 Peoples R China

This paper presents a reinforcement learning (RL)-based energy management strategy for a hybrid electric tracked vehicle. A control-oriented model of the powertrain and vehicle dynamics is first established. According to the sample information of the experimental driving schedule, statistical characteristics at various velocities are determined by extracting the transition probability matrix of the power request. Two RL-based algorithms, namely q-learning and Dyna algorithms, are applied to generate optimal control solutions. The two algorithms are simulated on the same driving schedule, and the simulation results are compared to clarify the merits and demerits of these algorithms. Although the q-learning algorithm is faster (3 h) than the Dyna algorithm (7 h), its fuel consumption is 1.7% higher than that of the Dyna algorithm. Furthermore, the Dyna algorithm registers approximately the same fuel consumption as the dynamic programming-based global optimal solution. The computational cost of the Dyna algorithm is substantially lower than that of the stochastic dynamic programming.

关键词： reinforcement learning (RL) hybrid electric tracked vehicle (HETV) q-learning algorithm Dyna algorithm dynamic programming (DP) stochastic dynamic programming (SDP)

来源：评论

学校读者我要写书评

暂无评论

Joint Data Offloading and Energy-Efficient Secure MEC Resource Allocation Method for IoT Device Data in RAN Communication

IEEE TRANSACTIONS ON GREEN COMMUNICATIONS AND NETWORKING

引用

IEEE TRANSACTIONS ON GREEN COMMUNICATIONS AND NETWORKING 2024年第3期8卷 1008-1017页

作者： He, qiang Li, Ji Zhu, Xiaogang Jolfaei, Alireza Feng, Zheng Tolba, Amr Yu, Keping Fu, Yukai Northeastern Univ Coll Med & Biol Informat Engn Shenyang 110169 Peoples R China Nanchang Univ Sch Publ Policy & Adm Nanchang 330031 Jiangxi Peoples R China Jiangxi Inst Ind Technol Internet Things Sch Elect & Informat Engn Yingtan 335000 Jiangxi Peoples R China Flinders Univ S Australia Coll Sci & Engn Adelaide SA 5042 Australia King Saud Univ Community Coll Comp Sci Dept Riyadh 11437 Saudi Arabia Hosei Univ Grad Sch Sci & Engn Tokyo 1028160 Japan Chinese Acad Sci Shenyang Inst Automat Shenyang 110016 Peoples R China

As information technology rapidly advances, 5G technology, Radio Access Networks (RAN), and the Internet of Things (IoT) have emerged as the core elements of next-generation communication technology. There is an increasing demand for real-time communication and reduced latency in various applications. Therefore, this paper proposes a four-layer Mobile Edge Computing (MEC) architecture that connects user devices to the core network using RAN. Blockchain verification is used for data storage and access permission separation. The architecture aims to address the high latency, low flexibility, and security issues in cloud computing communication. We also propose a MEC server location algorithm to optimize communication distance, and a q-learning algorithm for selection and resource allocation. Experimental results demonstrate significant energy savings compared to baseline algorithms.

关键词： Servers Computer architecture Internet of Things Task analysis Energy consumption Security Optimization Mobile edge computing radio access networks resource allocation energy consumption q-learning algorithm

来源：评论

学校读者我要写书评

暂无评论

Structural and behavioural evaluation of renewable energy power plants' impacts on transmission network congestion using an analytical approach

引用

IET RENEWABLE POWER GENERATION 2020年第7期14卷 1164-1173页

作者： Poursalimi Jaghargh, Mohammad Javad Rajabi Mashhadi, Habib Ferdowsi Univ Mashhad FUM Dept Elect Engn Fac Engn Mashhad Razavi Khorasan Iran

Renewable energy support schemes implementation increases the significance of renewable energy power plants (REPPs) in studying power market characteristics. In this study, an analytical approach is presented to investigate the impact of REPPs on transmission lines congestion. For this purpose, considering REPPs, social welfare maximisation problem corresponding to the spot electricity market should be solved by the independent system operator. A closed-form expression of the Lagrange multiplier (LM) of each congested line is presented to analytically demonstrate the role of the power system structure and strategic behaviour of conventional generation companies (GenCos) on the network congestion. The structural part of LM expression is decomposed into four components to distinguish the impact of GenCos from the impact of REPPs. The GenCos' strategic behaviour in the spot market bidding problem is simulated using well-known q-learning algorithm. The results confirm that REPPs deployment can remarkably affect both the structural and behavioural parts of the LM and emphasise the significance of the GenCos strategic behaviour on the occurrence of congestion. The results illustrate that since renewable energy support schemes may increase the possibility of congestion and GenCos' strategic behaviour, it is essential to take power system characteristics into account in their design.

关键词： power generation economics renewable energy sources power markets learning (artificial intelligence) power engineering computing renewable energy support schemes implementation renewable energy power plants power market characteristics analytical approach transmission lines congestion social welfare maximisation problem spot electricity market independent system operator closed-form expression congested line power system structure LM expression GenCos' strategic behaviour spot market bidding problem REPPs deployment structural parts behavioural parts GenCos strategic behaviour power system characteristics structural evaluation behavioural evaluation transmission network congestion Lagrange multiplier q-learning algorithm

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Reinforcement learning and Parallel Computing Applied to the K-Server Problem

引用

IEEE LATIN AMERICA TRANSACTIONS 2016年第10期14卷 4351-4357页

作者： Costa, M. L. Padilha, C. A. A. Melo, J. D. Doria Neto, A. D. Univ Estado Rio Grande Norte UERN Mossoro RN Brazil Univ Estado Rio Grande Norte UERN Natal RN Brazil Univ Fed Rio Grande do Sul Inst Informat Porto Alegre RS Brazil

In this paper was proposed an algorithm based on Hierarchical Reinforcement learning (HRL) and Parallel Computing to solve an online computing problem, the K-Server Problem (KSP). The size of the storage structure used for reinforcement learning to obtain the optimal policy grows exponentially with the number of states and actions, limiting its use to smaller problems due to the curse of dimensionality. The problem is modeled as a multiple steps decision process computed in parallel by applying the q-learning algorithm to obtain optimal policies in a reduced number of nodes obtained from an clustering process. The results show the applicability of the proposed method to real problems of large size.

关键词： Metrical Task Systems The K-Server Problem Curse of Dimensionality Hierarchical Reinforcement learning q-learning algorithm Parallel Computing

来源：评论

学校读者我要写书评

暂无评论

Adaptive biasing scheme for load balancing in backhaul constrained small cell networks

引用

IET COMMUNICATIONS 2015年第7期9卷 999-1005页

作者： Xu, Yang Yin, Rui Yu, Guanding Zhejiang Univ Dept Informat Sci & Elect Engn Hangzhou 310027 Zhejiang Peoples R China Hangzhou Dianzi Univ Zhejiang Prov Key Lab Data Storage & Transmiss Te Hangzhou 310018 Zhejiang Peoples R China Zhejiang Gongshang Univ Coll Informat & Elect Engn Hangzhou 310018 Zhejiang Peoples R China

In this study, a distributed biasing scheme is designed to achieve load balancing for heterogeneous networks. Based on the limited backhaul capacity and user distribution in the system, each small cell base station adaptively and distributively changes its cell range by setting the bias value, to effectively utilise the wireless resource and achieve load balancing as well. The q-learning algorithm is adopted to design the biasing scheme in each small cell base station. The tradeoff between the backhaul resource utilisation and the quality-of-service of users is considered in the reward function of the q-learning model. To examine the performance of the distributed scheme, a centralised scheme aiming at maximising the backhaul resource utilisation is also proposed for comparison, whose performance lower bound is derived. Numerical results show that the proposed distributed scheme can effectively utilise the backhaul resource for load balancing, and achieve a close performance to the centralised one.

关键词： cellular radio resource allocation learning (artificial intelligence) quality of service numerical analysis load balancing backhaul constrained small cell networks adaptive distributed biasing scheme user distribution system small cell base station wireless resource q-learning algorithm backhaul resource utilisation quality-of-service numerical analysis

来源：评论

学校读者我要写书评

暂无评论

Reinforcement learning control with adaptive gain for a Saccharomyces cerevisiae fermentation process

引用

APPLIED SOFT COMPUTING 2011年第8期11卷 4488-4495页

作者： Li, Dazi qian, Li Jin, qibing Tan, Tianwei Beijing Univ Chem Technol Coll Informat Sci & Technol Dept Automat Beijing 100029 Peoples R China Beijing Univ Chem Technol Coll Life Sci & Technol Beijing Key Lab Bioproc Beijing 100029 Peoples R China

A tight and robust yeast fermentation controller is usually difficult to achieve because of the inherent uncertainty, nonlinear, and time-varying characteristics of the yeast fermentation dynamic process. This paper presented an alternative method for yeast fermentation process control by hybrid reinforcement learning algorithm and fuzzy logic. The fuzzy logic was used to adjust the weighting gain of control action adaptively from reinforcement learning. It led to faster tracking and helped to alleviate the overshoot of the controller. The improved multi-step action q-learning control algorithm was developed and demonstrated through studies on ethanol concentration control of the yeast fermentation process. Experimental results show that the improved multi-step action q-learning controller has much lower overshoot, faster tracking, shorter transition, and smoother control signal than the advanced PID controller. (C) 2011 Elsevier B.V. All rights reserved.

关键词： q-learning algorithm Multi-step action Fermentation process Adaptive gain

来源：评论

学校读者我要写书评

暂无评论

Models and algorithms for multi-crane oriented scheduling method in container terminals

引用

TRANSPORT POLICY 2009年第5期16卷 271-278页

作者： Zeng, qingcheng Yang, Zhongzhen Lai, Luyuan Dalian Maritime Univ Coll Transportat Management Dalian Peoples R China

"Multi-crane oriented" is a scheduling method that yard trailers can be shared by different quay cranes. In this paper, two models for this problem are developed. The first one is a model for an inter-ship-based sharing method. In this model, yard trailers can be shared by quay cranes of different ships. To solve the model, a two-phase Tabu search algorithm is designed. The second one is a model for a ship-based sharing method. In this model, yard trailers can only be shared by quay cranes of the same ship. q-learning algorithm is designed to solve the model. Numerical tests show that the "multi-crane oriented" method can decrease the yard trailers' travel distance, reduce the disequilibrium of different working lines, and thus improve the operation efficiency in container terminals. Crown Copyright (C) 2009 Published by Elsevier Ltd. All rights reserved.

关键词： Container terminals Multi-oriented scheduling method Yard trailer scheduling q-learning algorithm

来源：评论

学校读者我要写书评

暂无评论

Application of intelligent system based on deep reinforcement learning in electrical engineering automation control

引用

INTERNATIONAL JOURNAL OF GRID AND UTILITY COMPUTING 2024年第3-4期15卷 323-332页

作者： Wu, Zhihe Dalian Jiaotong Univ Automat & Elect Engn Dalian Liaoning Peoples R China

The application of intelligent control technology in the power electronics industry has promoted the development of power automation. It not only changes its control and management mode, but also greatly improves efficiency. However, in the power automation system, it is necessary to fully consider its use efficiency according to the actual situation, and gradually promote the application of intelligent technology in power automation technology. This paper focuses on analysing and discussing the application of intelligent technology in power system and provides reference for future work. On this basis, a method of applying deep reinforcement learning method to automatic control of power engineering is proposed. This paper introduces a learning algorithm based on artificial emotion augmentation to improve the operational performance of power grids. The relationship between artificial emotion and reinforcement learning in artificial psychology is discussed from three perspectives: behaviour value selection, q-value matrix update and reward value function update. According to the experimental and calculation, in the ACE simulation results, it can be known that the q-learning method, the q(lambda) method, and the DqL method are reduced by 39.7%, 55.8%, and 61.7%, respectively, in the Delta f simulation results, the q-learning algorithm, the q(lambda)-learning algorithm and the deep q-learning method are 58.3%, 75% and 75% lower than PID, respectively. Simulation experiments showed that the algorithm outperforms the other three algorithms.

关键词： electrical engineering automation control deep reinforcement learning smart system q-learning algorithm traditional power system

来源：评论

学校读者我要写书评

暂无评论

Intelligent Secure Communication for Cognitive Networks With Multiple Primary Transmit Power

引用

IEEE ACCESS 2020年 8卷 37343-37351页

作者： Lai, Shiwei Xia, Junjuan Zou, Dan Fan, Liseng Guangzhou Univ Sch Comp Sci & Cyber Engn Guangzhou 510006 Peoples R China East China Jiaotong Univ Sch Informat Engn Nanchang 330013 Jiangxi Peoples R China

In this paper, we study an intelligent secure communication scheme for cognitive networks with multiple primary transmit power, where a secondary Alice transmits its secrecy data to a secondary Bob threatened by a secondary attacker. The secondary nodes limit their transmit power among multiple levels, in order to maintain the quality of service of the primary networks. The attacker can work in an eavesdropping, spoofing, jamming or silent mode, which can be viewed as the action in the traditional q-learning algorithm. On the other hand, the system can adaptively choose the transmit power level among multiple ones to suppress the intelligent attacker, which can be viewed as the status of q-learning algorithm. Accordingly, we firstly formulate this secure communication problem as a static secure communication game with Nash equilibrium (NE) between the main links and attacker, and then employ the q-learning algorithm to select the transmit power level. Simulation results are finally demonstrated to verify that the intelligent attacker can be effectively suppressed by the proposed studies in this paper.

关键词： Intelligent secure communication q-learning algorithm Nash equilibrium

来源：评论

学校读者我要写书评

暂无评论

Dynamic Resource Allocation Strategy of Multi-Objective Fuzzy Optimization Based on Markov Decision Process

引用

IEEE ACCESS 2023年 11卷 99607-99613页

作者： Zhang, Yu-Ting Yang, Jing-Yu Wu, Yang Natl Def Univ Grad Sch Beijing 100091 Peoples R China Natl Def Univ Joint Operat Coll Beijing 100091 Peoples R China

With the limited amount of national defense resources, the dynamic allocation of resources in overall project construction is conducive to improving the efficiency of project construction. In this regard, this paper constructs a dynamic resource allocation model of multi-objective and multi-stage fuzzy optimization based on the Markov decision process (MDP). On this basis, this paper further adopts the q-learning algorithm to seek strategy optimization. At last, this paper determines the resource allocation strategy that can optimize the total benefit of construction projects through the example analysis. Relevant results reveal that the dynamic resource allocation model of multi-objective and multi-stage fuzzy optimization based on MDP can achieve the effect of optimizing the project construction objectives and maximizing the construction output.

关键词： Resource management Dynamic scheduling Optimization q-learning Heuristic algorithms Indexes Markov processes Fuzzy systems Process planning Decision making Dynamic resource allocation fuzzy optimization Markov decision process model 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：