基于融合进化算法的用户日负荷曲线聚类分析

Cluster Analysis of User Daily Load Curve Based on Fusion Evolutionary Algorithm

负荷分类对电网调度、负荷预测、用户用电行为分析等具有重要意义.针对传统负荷分类算法易陷入局部最优解而无法确定最优初始聚类中心,导致分类结果不准确问题,提出一种融合进化算法优化模糊C均值(FCM)的负荷聚类算法.首先使用重心Lagrange插值法填充负荷曲线缺失点,其次利用线性函数将不同行业负荷曲线归一化,最后结合遗传算法全局搜索效率高以及模拟退火算法计算时间短的特点优化FCM进行负荷聚类,弥补了传统FCM易陷入局部最优解的问题.算例表明:所提算法聚类中心距离较远,用户日负荷曲线分类结果较准确;相较于传统FCM不易陷入局部最优解,且具有一定的鲁棒性.

Load classification is of great significance to power grid dispatching,load forecasting,and user behavior analysis.Aiming at the problem that the traditional load classification algorithm is easy to fall into the local optimal solution and cannot determine the optimal initial clustering center,which causes the classification result to be inaccurate,a load clustering algorithm that optimizes the fuzzy C-means(FCM)by fusion evolution algorithm is proposed.First,the Lagrange interpolation method is used to fill in the missing points of the load curve,and then the linear function is used to normalize the load curves of different industries.Finally,combining the characteristics of high global search efficiency of genetic algorithm and short calculation time of simulated annealing algorithm,FCM is optimized for load clustering,which makes up for the problem that traditional FCM is easy to fall into local optimal solution.The calculation example shows that the proposed algorithm is far away from the clustering center,and the user’s daily load curve classification result is more accurate.Compared with traditional FCM,it is not easy to fall into a local optimal solution,and it has certain robustness.