模糊C均值聚类在蓄电池SOC预测中的应用  

Application of Fuzzy Clustering Method in Predicting Batteries' SOC

作　　者：周奇[1] 罗培[1] 

机构地区：[1]湘潭大学信息工程学院,湘潭411105

出　　处：《电源学报》2014年第4期99-104,共6页Journal of Power Supply

摘　　要：针对铅酸蓄电池数学模型难以建立以及荷电状态精确预测问题,本文提出了一种利用模糊C-均值聚类算法对蓄电池SOC控制器参数及结构进行辨识的建模方法。通过对蓄电池电动势、内阻等实时数据进行聚类分析,能够有效地实现输入空间划分并能够有针对性地生成模糊控制规则,在此基础上构建了完整的荷电状态预测控制器。通过铅酸蓄电池放电实验验证了该预测方法的有效性。Since it is difficult to establish a mathematical model of lead-acid battery and to predict the accuracy of SOC, the fuzzy C-means （FCM） for recognizing the parameters and structure of battery SOC controller was presented. By clustering the battery electromotive force and internal resistance were collected in real time, this controller is effective to partition the space and construct the fuzzy control rules and it also constructs a complete state of charge prediction controller based on it. The battery discharge test verifies the effectiveness of the prediction method finally.

关 键 词：模糊C-均值聚类 铅酸蓄电池 控制规则 荷电状态 

分 类 号：TM912.1[电气工程—电力电子与电力传动]