非线性季节型电力负荷曲线集成聚类算法研究
|
摘要
传统方法直接设定聚类数量,得到的结果并非最优聚类数,且针对大规模电力数据,单一聚类方法无法同时达到聚类精度与效率两方面要求。为此,提出一种新的非线性季节型电力负荷曲线集成聚类算法。用DBI指标对聚类效果进行评价,将与DBI最小值相应的聚类数据作为最优聚类数量。通过模糊C均值聚类方法,依据原始非线性季节型电力数据对象间的相似程度,获取初始聚类中心。利用聚类性能更优、稳定性更高的层次聚类方法完成对聚类中心的组合,获取有效的集成聚类结果。实验结果表明,所提方法能够同时保证聚类精度与聚类效率,整体性能较强。
Traditional clustering methods directly set the number of clusters, and the result is not the optimal number of clusters. For large-scale power data, a single clustering method can not meet the requirements of clustering accuracy and efficiency at the same time. Therefore, a new integrated clustering algorithm for nonlinear seasonal power load curve is proposed. The clustering effect is evaluated by DBI index, and the clustering data corresponding to the minimum DBI is taken as the optimal clustering number. Based on the similarity between the original nonlinear seasonal power data objects, the initial clustering center is obtained by using the fuzzy C-means clustering method. The hierarchical clustering method with better clustering performance and higher stability is used to complete the combination of clustering centers and obtain effective integrated clustering results. The experimental results show that the proposed method can guarantee both clustering accuracy and clustering efficiency, and the overall performance is strong.
Traditional clustering methods directly set the number of clusters, and the result is not the optimal number of clusters. For large-scale power data, a single clustering method can not meet the requirements of clustering accuracy and efficiency at the same time. Therefore, a new integrated clustering algorithm for nonlinear seasonal power load curve is proposed. The clustering effect is evaluated by DBI index, and the clustering data corresponding to the minimum DBI is taken as the optimal clustering number. Based on the similarity between the original nonlinear seasonal power data objects, the initial clustering center is obtained by using the fuzzy C-means clustering method. The hierarchical clustering method with better clustering performance and higher stability is used to complete the combination of clustering centers and obtain effective integrated clustering results. The experimental results show that the proposed method can guarantee both clustering accuracy and clustering efficiency, and the overall performance is strong.
引文
[1] 张斌,庄池杰,胡军,等.结合降维技术的电力负荷曲线集成聚类算法.中国电机工程学报,2015;35(15):3741-3749.
[2] 宋易阳,李存斌,祁之强.基于云模型和模糊聚类的电力负荷模式提取方法.电网技术,2014;38(12):3378-3383.
[3] 祖向荣,田敏,白焰.基于模糊聚类与函数小波核回归的短期负荷预测方法.电力自动化设备,2016;36(10):134-140.
[4] 徐源,程潜善,李阳,等.基于大数据聚类的电力系统中长期负荷预测[J].电力系统及其自动化学报,2017,29(8):43-48.
[5] 张苑农,董静.非线性电子网络传输功率稳定性控制仿真[J].计算机仿真,2017,34(2):334-337.
[6] 刘正春,朱长青,赵锦成,等.独立电力系统非线性负荷暂态特性仿真[J].高电压技术,2017,43(1):329-336.
[7] 赵硕,张少敏.分布式电力负荷预测算法研究.小型微型计算机系统,2014;35(8):1856-1860.
[8] 白杨,王鹏,韩肖清,等.基于负荷不确定性建模的静态电压稳定性风险评估[J].中国电机工程学报,2016,36(13):3470-3478.
[9] 黄媛玉,毛弋,娄宁娜.基于Kohonen聚类小波包分解和ELM的短期电力负荷预测[J].湖南师范大学自然科学学报,2016,39(4):53-58.
[10] 彭勃,张逸,熊军,等.结合负荷形态指标的电力负荷曲线两步聚类算法.电力建设,2016;37(6):96-102.
[2] 宋易阳,李存斌,祁之强.基于云模型和模糊聚类的电力负荷模式提取方法.电网技术,2014;38(12):3378-3383.
[3] 祖向荣,田敏,白焰.基于模糊聚类与函数小波核回归的短期负荷预测方法.电力自动化设备,2016;36(10):134-140.
[4] 徐源,程潜善,李阳,等.基于大数据聚类的电力系统中长期负荷预测[J].电力系统及其自动化学报,2017,29(8):43-48.
[5] 张苑农,董静.非线性电子网络传输功率稳定性控制仿真[J].计算机仿真,2017,34(2):334-337.
[6] 刘正春,朱长青,赵锦成,等.独立电力系统非线性负荷暂态特性仿真[J].高电压技术,2017,43(1):329-336.
[7] 赵硕,张少敏.分布式电力负荷预测算法研究.小型微型计算机系统,2014;35(8):1856-1860.
[8] 白杨,王鹏,韩肖清,等.基于负荷不确定性建模的静态电压稳定性风险评估[J].中国电机工程学报,2016,36(13):3470-3478.
[9] 黄媛玉,毛弋,娄宁娜.基于Kohonen聚类小波包分解和ELM的短期电力负荷预测[J].湖南师范大学自然科学学报,2016,39(4):53-58.
[10] 彭勃,张逸,熊军,等.结合负荷形态指标的电力负荷曲线两步聚类算法.电力建设,2016;37(6):96-102.