摘要
着重论述了电力负荷预测中建模变量的选择、数据的预处理方法、模型的拓扑结构及其对预测精度的影响。针对水电企业电力负荷预测,提出了一种将经典的AR模型与T S模糊神经系统相结合的负荷预测方法。应用该方法对某水电厂近两年发电量作了预测计算,并与实测数据作了比较,表明该方法具有较好的鲁棒性和较高的精度。
The variable selection, data preprocessing and model structure of power load forecasting, and their effects on forecasting precision are discussed. A power load forecasting method which combines classical AR(Auto Regression) model with TS fuzzyneural system is put forward especially for power dam. The recent two years' loads of a power dam are forecasted with it. Compared the forecasted loads with the true loads, it shows that this method has better robusticity and higher precision.
出处
《电力自动化设备》
EI
CSCD
北大核心
2002年第10期43-46,共4页
Electric Power Automation Equipment
