改进多目标萤火虫优化的软子空间聚类算法及在短期负荷预测中的应用  

SOFT SUBSPACE CLUSTERING ALGORITHM BASED ON IMPROVED MULTI-OBJECTIVE FIREFLY AND ITS APPLICATION IN SHORT TERM LOAD FORECASTING

作　　者：张曦 康平 付雪峰[1,2] 叶军[1,2] 赵嘉[1,2,3] Zhang Xi;Kang Ping;Fu Xuefeng;Ye Jun;Zhao Jia(School of Information Engineering,Nanchang Institute of Technology,Nanchang 330000,Jiangxi,China;Jiangxi Province Key Laboratory of Water Information Cooperative Sensing and Intelligent Processing,Nanchang Institute of Technology,Nanchang 330000,Jiangxi,China;National-Local Engineering Laboratory of Water Engineering Safety and Effective Utilization of Resources in Poyang Lake Area,Nanchang Institute of Technology,Nanchang 330000,Jiangxi,China)

机构地区：[1]南昌工程学院信息工程学院,江西南昌330000 [2]南昌工程学院江西省水信息协同感知与智能处理重点实验室,江西南昌330000 [3]南昌工程学院鄱阳湖流域水工程安全与资源高效利用国家地方联合工程实验室,江西南昌330000

出　　处：《计算机应用与软件》2022年第7期261-268,321,共9页Computer Applications and Software

基　　金：国家自然科学基金项目(51669014);江西省杰出青年基金项目(2018ACB21029);江西省自然科学基金项目(20192BAB207031)。

摘　　要：针对传统软子空间聚类算法因单目标优化无法准确聚类的问题,提出一种改进多目标萤火虫优化的软子空间聚类算法(IMOFASSC)。对多目标萤火虫算法的步长因子和初始吸引力进行动态定义以弥补算法易提前收敛的缺陷,并设计一种萤火虫单行随机学习机制来提高最优解集分布的均匀性;将改进的多目标萤火虫算法运用到软子空间聚类问题中,同时优化簇内紧凑度、簇间分离度及负权值熵三个目标函数,将IMOFASSC应用到短期负荷预测中。实验结果表明,IMOFASSC不仅在低维和高维数据聚类中有较好的聚类效果,而且在短期负荷预测中具有推广应用价值。To solve the problem that the traditional soft subspace clustering algorithms cannot accurately cluster because of single objective optimization, this paper proposes a soft subspace clustering algorithm based on improved multi-objective firefly(IMOFASSC). We dynamically defined the step factor and initial attractiveness of multi-objective firefly algorithm to make up for the problem that algorithm was prone to premature convergence. We designed a firefly single-row random learning mechanism, so as to improve the uniformity of the distribution of the optimal solution set. The improved multi-objective firefly algorithm was applied to soft subspace clustering problem. And we optimized three objective functions, which were within-class compactness, between-class separation and negative weight entropy. IMOFASSC was applied in short term load forecasting. The experimental results show that IMOFASSC has better clustering effect in low-dimensional and high-dimensional data clustering, and has popular application value in short term load forecasting.

关 键 词：高维数据聚类 软子空间聚类算法 多目标优化问题 多目标萤火虫算法 短期负荷预测 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]