The Application and Research of the PAR Approach in the Short Term Load Forecasting

摘要

For load sequences changes run in cycle by days, weeks, and years, and the power load data are non-stationary, the periodic autoregressive (PAR) model is used to describe the periodic variations accurately of the power load and establish a short-term forecast of the prediction model. Compared with traditional time series, it is show that this way is more effective.