基于改进PSO-Kmeans算法的实际日负荷曲线聚类分析

Cluster Analysis of Actual Daily Load Curve Based on Improved PSO-Kmeans Algorithm

有效的实时用户日负荷曲线分类可为电网系统规划、负荷建模、负荷预测及需求侧管理等方面提供依据,同时为电网工作人员实现对负荷模型的判别提供帮助。为此提出一种基于改进PSO进化算法优化K-means的日负荷曲线用户行业分类方法。首先利用牛顿插值法对缺失数据进行填补,然后运用定值线性函数对数据归一化,最后采用权重线性递减及同步学习因子改进PSO解决算法易陷入局部最优解的问题,以优化K-means分类结果准确性。算例分析表明,PSO-Kmeans算法迭代能力强,具有全局寻优能力,且具有一定的鲁棒性,相较于传统K-means分类准确率高。

Effective real-time user daily load curve classification provides a basis for grid system planning,load modeling,load forecasting,and demand side management,and helps the power grid staff to distinguish the load model.An industry classification method based on K-means optimized by PSO evolutionary algorithm is proposed in this paper.Firstly,Newton interpolation method is used to fill in the missing data,and then constant value linear function is used to normalize the data.Finally,the PSO algorithm is improved by linear decreasing weight and synchronous learning factor to solve the problem that the algorithm is easy to fall into local optimal solution,so as to optimize the accuracy of K-means classification results.The example analysis shows that PSO-Kmeans algorithm has strong iterative ability,global optimization ability and certain robustness,and has higher classification accuracy than traditional K-means algorithm.