摘要
灰色预测模型GM(1,1)主要适合于光滑数据序列的预测,对非光滑数据序列常采用对数变换法、开n次方变换法和指数加权变换法提高数据的光滑度.由于月电量呈现明显的季节性变换,常用的提高光滑度的方法效果并不明显.文章提出了一种基于季节指数和灰色预测的月电量预测模型,以南京市某供电分公司近五年的月电量数据为基础,采用季节指数变换的灰色预测模型进行预测相比较对数变换法的灰色预测模型,预测效果得到了明显提高.
The grey forecast model GM （1, 1） fits in with the forecast of smooth data. With regard to non-smooth data, it adopts the logarithm transformation, the extraction of a root and the exponent weighted to improve their smoothness. Owing to the fact that the data of monthly electric power demand change in every season, the above-mentioned methods prove to be inadequate. To solve the problem, the paper puts forward a forecast model based on season exponent and grey forecast. Take the electric power demand data from 2000 to 2004 of a branch power supply company located in Nanjing as a sample, the results of forecast by this model show that it is more valid than the previous ones.
出处
《南京工程学院学报（自然科学版）》
2006年第1期1-6,共6页
Journal of Nanjing Institute of Technology(Natural Science Edition)
基金
江苏省教育厅高校哲学社会科学研究项目(03SJD630044)
关键词
季节指数
灰色预测
月电量
season exponent
grey forecast
monthly electric power demand
