摘要
电力负荷数据具有很强的非线性和随机性。为了提高负荷预测模型的精度,提出了一种基于聚类分析与神经网络相结合的预测模型。采用K均值聚类分析算法对影响负荷的各种因素进行分类处理,再选择预测日所属类别的历史数据作为训练样本对BP神经网络模型进行建模;利用MATLAB对北京市朝阳区的负荷量进行预测分析。结果表明未聚类预测模型的平均相对误差为6.4633%,聚类后的负荷预测模型平均相对误差为2.1431%。可见对历史数据进行聚类后建立的负荷预测模型误差更小,预测精度更高。
Power load data has strong nonlinearity and randomness. In order to improve the accuracy of load forecasting model,a prediction model based on cluster analysis and neural network is proposed. Firstly,the K-means clustering analysis algorithm is used to classify various factors that affect the load,and the historical data of the category of the forecast day is selected as the training sample to model the BP neural network. Finally,using MATLAB to predict the load in Chaoyang District of Beijing,the results show that the average relative error of the unclustered prediction model is 6. 4633%,and the average relative error of the clustered load forecasting model is 2. 1431%. It can be seen that the load forecasting model established after clustering the historical data has smaller errors and higher prediction accuracy.
作者
刘田梦
王丽婕
马嫒
LIU Tianmeng;WANG Lijie;MA Ai(School of Automation,Beijing Information Science & Technology Univemity,Beijing 100192,China;Beijing Aerospace Times Optical-electronic Technology Co.Ltd,Beijing 100092,China)
出处
《北京信息科技大学学报(自然科学版)》
2018年第4期24-28,共5页
Journal of Beijing Information Science and Technology University
基金
国家自然科学基金资助项目(51607009)
关键词
电力系统
短期负荷预测
K均值聚类分析
BP神经网络
electric power system
short-term load forecasting
cluster analysis
BP neural network