| 摘要: |
| 针对短期负荷预测方法中传统的模糊C均值（FCM）聚类容易陷入局部最优和对初始聚类中心敏感的问题，提出利用粒子群优化（PSO）算法的全局搜索特性来优化此缺点.通过优化的FCM聚类来选取与预测日相似的日期作为支持向量机的训练样本，既强化了训练样本的数据规律，又保证数据特征的一致性.实验结果表明，优化预测模型的预测精度优于BP神经网络和支持向量机算法. |
| 关键词: 短期负荷预测 相似日 相似性 模糊C均值(FCM)聚类 粒子群优化(PSO)算法 支持向量机(SVM) |
| DOI：10.3969/J.ISSN.1000-5137.2017.04.016 |
| 分类号:TP315.69 |
| 基金项目: |
|
| Gas load forecasting based on optimized fuzzy c-mean clustering analysis of selecting similar days |
|
Qiu Jing, Xu Xiaozhong, Deng Song, Wang Ting
|
|
The College of Information, Mechanical and Electrical Engineering, Shanghai Normal University, Shanghai 200234, China
|
| Abstract: |
| Traditional fuzzy c-means (FCM) clustering in short term load forecasting method is easy to fall into local optimum and is sensitive to the initial cluster center.In this paper,we propose to use global search feature of particle swarm optimization (PSO) algorithm to avoid these shortcomings,and to use FCM optimization to select similar date of forecast as training sample of support vector machines.This will not only strengthen the data rule of training samples,but also ensure the consistency of data characteristics.Experimental results show that the prediction accuracy of this prediction model is better than that of BP neural network and support vector machine (SVM) algorithms. |
| Key words: short term load forecasting similar days similarity fuzzy c-means (FCM) clustering particle swarm optimization (PSO) algorithm support vector machine (SVM) |
