On the process of power system black start after an accident, it can help to optimize the resources allocation and accelerate the recovery process that decomposing the power system into several independent partitions ...On the process of power system black start after an accident, it can help to optimize the resources allocation and accelerate the recovery process that decomposing the power system into several independent partitions for parallel recovery. On the basis of adequate consideration of fuzziness of black-start zone partitioning, a new algorithm based on fuzzy clustering analysis is presented. Characteristic indexes are extracted fully and accurately. The raw data matrix is made up of the electrical distance between every nodes and blackstart resources. Closure transfer method is utilized to get the dynamic clustering. The availability and feasibility of the proposed algorithm are verified on the New-England 39 bus system at last.展开更多
Water quality assessment of lakes is important to determine functional zones of water use.Considering the fuzziness during the partitioning process for lake water quality in an arid area,a multiplex model of fuzzy clu...Water quality assessment of lakes is important to determine functional zones of water use.Considering the fuzziness during the partitioning process for lake water quality in an arid area,a multiplex model of fuzzy clustering with pattern recognition was developed by integrating transitive closure method,ISODATA algorithm in fuzzy clustering and fuzzy pattern recognition.The model was applied to partition the Ulansuhai Lake,a typical shallow lake in arid climate zone in the west part of Inner Mongolia,China and grade the condition of water quality divisions.The results showed that the partition well matched the real conditions of the lake,and the method has been proved accurate in the application.展开更多
随着高比例、大规模分布式光伏并网以及电动汽车的普及,如何发挥电动汽车灵活性、实现配电网分布式光伏与本地电动汽车负荷灵活性资源的友好协调是当前需要解决的重要问题。为此,提出了考虑电动汽车与分布式光伏协同的配电网集群划分与...随着高比例、大规模分布式光伏并网以及电动汽车的普及,如何发挥电动汽车灵活性、实现配电网分布式光伏与本地电动汽车负荷灵活性资源的友好协调是当前需要解决的重要问题。为此,提出了考虑电动汽车与分布式光伏协同的配电网集群划分与运行策略。首先,建立电动汽车可调充电功率灵活性聚合模型,提出基于Louvain算法的改进模块度指标配电网分布式集群划分方法;其次,基于历史数据信息生成电动汽车多时间尺度充电场景,提出考虑电动汽车充电灵活性的分布式集群协同优化模型;最后,采用同步交替方向乘子法(synchronous alternating direction multiplier method,SADMM)实现各集群优化模型的分布式求解。仿真结果表明,利用电动汽车充电灵活性参与配电网协同运行可有效提高分布式光伏利用率,并且在满足电动汽车用户充电需求的同时保证了配电网电压运行安全。展开更多
针对复杂居民地多边形的信息挖掘问题,提出了一种多级图划分聚类分析方法,构造居民地多边形的图模型,并通过对图模型进行粗化匹配与重构、初始化分和细化得到聚类结果。首先构建研究区域内居民地建筑物的Delaunay三角网,生成包含研究对...针对复杂居民地多边形的信息挖掘问题,提出了一种多级图划分聚类分析方法,构造居民地多边形的图模型,并通过对图模型进行粗化匹配与重构、初始化分和细化得到聚类结果。首先构建研究区域内居民地建筑物的Delaunay三角网,生成包含研究对象之间的邻接信息图;然后结合空间认知准则和人类认知的特点,采用形状狭长度、面积比、凹凸性、距离和连通性5个指标度量邻接图的相似性;最后应用多级图划分方法,得到聚类结果。采用中国上海地区的居民地建筑物矢量数据进行聚类分析实验,并对比了改进的k均值算法(k-Means++)、具有噪声鲁棒性的基于密度的空间聚类算法(density-based spatial clustering of applications with noise,DBSCAN)和最小生成树(minimum spanning tree, MST)聚类算法得到的轮廓系数以及视觉效果。实验结果表明,基于多级图划分的居民地多边形聚类分析的结果更加符合人类认知。展开更多
文摘On the process of power system black start after an accident, it can help to optimize the resources allocation and accelerate the recovery process that decomposing the power system into several independent partitions for parallel recovery. On the basis of adequate consideration of fuzziness of black-start zone partitioning, a new algorithm based on fuzzy clustering analysis is presented. Characteristic indexes are extracted fully and accurately. The raw data matrix is made up of the electrical distance between every nodes and blackstart resources. Closure transfer method is utilized to get the dynamic clustering. The availability and feasibility of the proposed algorithm are verified on the New-England 39 bus system at last.
基金Supported by the National Natural Science Foundation of China (No.50269001, 50569002, 50669004)Natural Science Foundation of Inner Mongolia (No.200208020512, 200711020604)The Key Scientific and Technologic Project of the 10th Five-Year Plan of Inner Mongolia (No.20010103)
文摘Water quality assessment of lakes is important to determine functional zones of water use.Considering the fuzziness during the partitioning process for lake water quality in an arid area,a multiplex model of fuzzy clustering with pattern recognition was developed by integrating transitive closure method,ISODATA algorithm in fuzzy clustering and fuzzy pattern recognition.The model was applied to partition the Ulansuhai Lake,a typical shallow lake in arid climate zone in the west part of Inner Mongolia,China and grade the condition of water quality divisions.The results showed that the partition well matched the real conditions of the lake,and the method has been proved accurate in the application.
文摘随着高比例、大规模分布式光伏并网以及电动汽车的普及,如何发挥电动汽车灵活性、实现配电网分布式光伏与本地电动汽车负荷灵活性资源的友好协调是当前需要解决的重要问题。为此,提出了考虑电动汽车与分布式光伏协同的配电网集群划分与运行策略。首先,建立电动汽车可调充电功率灵活性聚合模型,提出基于Louvain算法的改进模块度指标配电网分布式集群划分方法;其次,基于历史数据信息生成电动汽车多时间尺度充电场景,提出考虑电动汽车充电灵活性的分布式集群协同优化模型;最后,采用同步交替方向乘子法(synchronous alternating direction multiplier method,SADMM)实现各集群优化模型的分布式求解。仿真结果表明,利用电动汽车充电灵活性参与配电网协同运行可有效提高分布式光伏利用率,并且在满足电动汽车用户充电需求的同时保证了配电网电压运行安全。
文摘针对复杂居民地多边形的信息挖掘问题,提出了一种多级图划分聚类分析方法,构造居民地多边形的图模型,并通过对图模型进行粗化匹配与重构、初始化分和细化得到聚类结果。首先构建研究区域内居民地建筑物的Delaunay三角网,生成包含研究对象之间的邻接信息图;然后结合空间认知准则和人类认知的特点,采用形状狭长度、面积比、凹凸性、距离和连通性5个指标度量邻接图的相似性;最后应用多级图划分方法,得到聚类结果。采用中国上海地区的居民地建筑物矢量数据进行聚类分析实验,并对比了改进的k均值算法(k-Means++)、具有噪声鲁棒性的基于密度的空间聚类算法(density-based spatial clustering of applications with noise,DBSCAN)和最小生成树(minimum spanning tree, MST)聚类算法得到的轮廓系数以及视觉效果。实验结果表明,基于多级图划分的居民地多边形聚类分析的结果更加符合人类认知。
