This research develops a comprehensive method to solve a combinatorial problem consisting of distribution system reconfiguration, capacitor allocation, and renewable energy resources sizing and siting simultaneously a...This research develops a comprehensive method to solve a combinatorial problem consisting of distribution system reconfiguration, capacitor allocation, and renewable energy resources sizing and siting simultaneously and to improve power system's accountability and system performance parameters. Due to finding solution which is closer to realistic characteristics, load forecasting, market price errors and the uncertainties related to the variable output power of wind based DG units are put in consideration. This work employs NSGA-II accompanied by the fuzzy set theory to solve the aforementioned multi-objective problem. The proposed scheme finally leads to a solution with a minimum voltage deviation, a maximum voltage stability, lower amount of pollutant and lower cost. The cost includes the installation costs of new equipment, reconfiguration costs, power loss cost, reliability cost, cost of energy purchased from power market, upgrade costs of lines and operation and maintenance costs of DGs. Therefore, the proposed methodology improves power quality, reliability and security in lower costs besides its preserve, with the operational indices of power distribution networks in acceptable level. To validate the proposed methodology's usefulness, it was applied on the IEEE 33-bus distribution system then the outcomes were compared with initial configuration.展开更多
能源互联网现行调控模式主要面向大负荷、大火电机组等能量大户,不适应其分布式能源资源(distributed energy resources,DER)渗透率不断提升的趋势。该文旨在建立多DER主体群智调控框架,通过在虚拟空间系统性地揭示并利用DER的聚合涌现...能源互联网现行调控模式主要面向大负荷、大火电机组等能量大户,不适应其分布式能源资源(distributed energy resources,DER)渗透率不断提升的趋势。该文旨在建立多DER主体群智调控框架,通过在虚拟空间系统性地揭示并利用DER的聚合涌现规律,激发其主观能动性,从而开启调度新模式。具体而言,拟以系统论、数据密集型科学发现范式(第四范式)等为指导思想,以虚拟孪生、大数据分析、机器学习与人机混合智能等为内核,以数字孪生、虚拟仿真推演、高维统计、时空数据分析、深度神经网络、人在回路与知识嵌入等为技术手段,设计并逐步完善“虚拟孪生+数据科学+系统论+第四范式”的系统性框架。该框架旨在通过数据贯通、数业融合、虚实交互等手段实现数据赋能提智工程系统,最终形成复杂系统调度新理论。展开更多
针对OpenStack云计算平台默认调度算法存在资源利用率低和负载不均衡的问题,提出一种基于改进的秃鹰搜索的调度算法(PieceWise bald and t-distribution eagle search,PBES),旨在最大化云数据中心的资源利用率和负载均衡。采用PieceWis...针对OpenStack云计算平台默认调度算法存在资源利用率低和负载不均衡的问题,提出一种基于改进的秃鹰搜索的调度算法(PieceWise bald and t-distribution eagle search,PBES),旨在最大化云数据中心的资源利用率和负载均衡。采用PieceWise混沌映射提高搜索算法的收敛速度和精度,引入t分布避免算法陷入局部最优解。综合考虑CPU、内存、磁盘和带宽等4种资源指标,采集真实环境下的数据并进行实验,其结果表明,相较OpenStack默认调度算法和粒子群算法,PBES算法在资源利用率和负载均衡方面都有显著提升。展开更多
文摘This research develops a comprehensive method to solve a combinatorial problem consisting of distribution system reconfiguration, capacitor allocation, and renewable energy resources sizing and siting simultaneously and to improve power system's accountability and system performance parameters. Due to finding solution which is closer to realistic characteristics, load forecasting, market price errors and the uncertainties related to the variable output power of wind based DG units are put in consideration. This work employs NSGA-II accompanied by the fuzzy set theory to solve the aforementioned multi-objective problem. The proposed scheme finally leads to a solution with a minimum voltage deviation, a maximum voltage stability, lower amount of pollutant and lower cost. The cost includes the installation costs of new equipment, reconfiguration costs, power loss cost, reliability cost, cost of energy purchased from power market, upgrade costs of lines and operation and maintenance costs of DGs. Therefore, the proposed methodology improves power quality, reliability and security in lower costs besides its preserve, with the operational indices of power distribution networks in acceptable level. To validate the proposed methodology's usefulness, it was applied on the IEEE 33-bus distribution system then the outcomes were compared with initial configuration.
文摘能源互联网现行调控模式主要面向大负荷、大火电机组等能量大户,不适应其分布式能源资源(distributed energy resources,DER)渗透率不断提升的趋势。该文旨在建立多DER主体群智调控框架,通过在虚拟空间系统性地揭示并利用DER的聚合涌现规律,激发其主观能动性,从而开启调度新模式。具体而言,拟以系统论、数据密集型科学发现范式(第四范式)等为指导思想,以虚拟孪生、大数据分析、机器学习与人机混合智能等为内核,以数字孪生、虚拟仿真推演、高维统计、时空数据分析、深度神经网络、人在回路与知识嵌入等为技术手段,设计并逐步完善“虚拟孪生+数据科学+系统论+第四范式”的系统性框架。该框架旨在通过数据贯通、数业融合、虚实交互等手段实现数据赋能提智工程系统,最终形成复杂系统调度新理论。
文摘针对OpenStack云计算平台默认调度算法存在资源利用率低和负载不均衡的问题,提出一种基于改进的秃鹰搜索的调度算法(PieceWise bald and t-distribution eagle search,PBES),旨在最大化云数据中心的资源利用率和负载均衡。采用PieceWise混沌映射提高搜索算法的收敛速度和精度,引入t分布避免算法陷入局部最优解。综合考虑CPU、内存、磁盘和带宽等4种资源指标,采集真实环境下的数据并进行实验,其结果表明,相较OpenStack默认调度算法和粒子群算法,PBES算法在资源利用率和负载均衡方面都有显著提升。
