以智慧城市为应用的基于智能体的分布式仿真

发布时间：2021-10-07 19:23

　　随着计算机仿真技术的发展和在其他学科中的应用,越来越多研究领域开始采用计算机仿真来作为重要的研究手段之一。在计算仿真技术中,分布式离散事件仿真随着分布式计算的进步得到了越来越多的应用。这类仿真方式对大规模的仿真应用具有较好的性能表现。另一方面,近年来,为了解决粮食短缺问题,在城市规划学研究中,城市农业逐渐成为受到人们关注的新兴概念。在城市农业中,食品生产消费链不再严格按照从农村到城市的单向结构,而是将城市与农业生产混合,通过前沿技术提高农业生产效率。本文的研究内容主要包含两个方面。首先是实现一种多智能体系统分布式仿真引擎的动态计算资源管理机制;其次是在城市农业系统情境下构建出一个智慧城市多智能体模型,并结合上述机制对其进行仿真。本研究的主要贡献列举如下:第一,对分布式离散事件仿真平台PDES-MAS设计了动态计算资源调整机制。我们从底层实现的角度对该机制进行了详细的描述。它在仿真执行过程中对系统中的计算资源进行监测。当发现因为仿真模型行为方式的变化而导致系统中出现计算资源不足,或者计算资源冗余的情况时,该机制能够动态地分配或回收计算资源。第二,针对分布式离散事件仿真中事件消息路由功能和... 

【文章来源】：哈尔滨工业大学黑龙江省 211工程院校 985工程院校

【文章页数】：99 页

【学位级别】：硕士

【文章目录】：
摘要
Abstract
Chapter 1 Introduction
    1.1 Background, Purpose and Significance of this Research
        1.1.1 Background of Multi-Agent Simulation
        1.1.2 Developmental of Intelligent Agriculture
        1.1.3 Purpose and Significance
    1.2 Related Work
        1.2.1 Research status of Parallel and Discrete Event Simulation
        1.2.2 Research Status of Intelligent Agriculture
    1.3 Main Research Content
        1.3.1 Multi-agent systems
        1.3.2 PDES Simulation Technology
        1.3.3 PDES-MAS Simulation Platform
        1.3.4 Load Balancing Algorithm
        1.3.5 Tile World Model
        1.3.6 Developmental of Intelligent Agriculture
        1.3.7 Validation of This Research
        1.3.8 Evaluation Index of This Research
    1.4 Research Framework and Methods
    1.5 Main Points of Innovation
    1.6 Organization of this Thesis
Chapter 2 Research on the Framework of Dynamic Capacity Adjustment
    2.1 Introduction
    2.2 PDES-MAS Architecture and Its Problems
    2.3 Limitations of Load Balancing Mechanism
    2.4 Research on the Implementation of Dynamic Capacity Expansion
        2.4.1 Node spawning method based on MPI standard
        2.4.2 Communicator Switching through Multithreading
        2.4.3 Deadlock Problem and Root-CLP Based Scheduling
        2.4.4 Criterion for Starting Node Spawning
    2.5 Research on the Implementation of Dynamic Capacity Contraction
        2.5.1 Criterion for Starting Node Deletion
        2.5.2 Implementation of Node Deletion
    2.6 Brief Summary
Chapter 3 Research on Dynamic Topology Adjustment Supporting MessageRouting
    3.1 Introduction
    3.2 Basic Concept of Message Routing Mechanism
    3.3 Effect of Load balancing Algorithm on Message Routing
    3.4 Implementation of Message Routing in PDES -MAS
    3.5 Design of Dynamic Capacity Adjustment for Message Routing
        3.5.1 Message Type Design
        3.5.2 Design of Dynamic Capacity Expansion for Message Routing
        3.5.3 Design of Dynamic Capacity Contraction for Message Routing
    3.6 Brief Summary
Chapter 4 Research on Dynamic Topology Adjustment Supporting RangeQuery
    4.1 Introduction
    4.2 Basic Concept of Range Query
    4.3 Implementation of Range Query in PDES -MAS
        4.3.1 Port Ranges of CLP nodes
        4.3.2 Range Period and Range Period List
    4.4 Design of Dynamic Adjustment Mechanism for Range Query
        4.4.1 Design of Dynamic Capacity Expansion for Range Query
        4.4.2 Design of Dynamic Capacity Contraction for Range Query
    4.5 Brief Summary
Chapter 5 Experiments on the Tile World Benchmark
    5.1 Introduction
    5.2 Experiment environment
    5.3 Implementation of Tile World model on PDES -MAS
    5.4 Simulation Experiment for Node Spawning
        5.4.1 Analysis of Simulation Conditions
        5.4.2 Comparison of Simulation Results
    5.5 Simulation Experiment for Node Deletion
        5.5.1 Analysis of Simulation Conditions
        5.5.2 Comparison of Simulation Results
    5.6 Brief summary
Chapter 6 Modeling and Simulation for Smart City Application
    6.1 Analysis of the Smart City System
    6.2 Implementation of Smart City Model on PDES -MAS
    6.3 Experiments on the Smart City Model
        6.3.1 Simulation Experiment for Node Spawning
        6.3.2 Simulation Experiment for Node Deletion
    6.4 Brief Summary
Conclusions
结论
References
Papers Published in the Period of Master Education
Acknowledgements



本文编号：3422629