Monthly electricity consumption prediction model based on grey theory and seasonal index model
Traditional seasonal index prediction model uses average value as seasonal index,so there exists inaccuracy problem.In order to predict monthly electricity consumption accurately and easily,introduces non-linear trend function based on traditional seasonal index algorithm and grey prediction model GM(1,1),aiming to better predict randomness.Improved seasonal index algorithm is produced.Through separate calculation of 44 groups of data in 4 periods in China's first sector and second sector,effectiveness of the improved algorithm is proved.Calculation results show that,compared with traditional seasonal index algorithm and traditional GM(1,1),the improved algorithm's MAPE reduces 2%～21%and 4%～10%,and the improved algorithm's RMSE reduces 39%～67%and 22%～76%.To sum up,the improved algorithm is an effective and simple method.
grey theoryseasonal index modelmonthly electricity consumptionprediction modelMAPERMSE
NETL
NSTL
万方数据
