Research of Wind Power Prediction Based on the Auto-Regressive Model

This paper discusses in detail the reason for the inaccurate result from the present system for wind farm output power prediction. Time series analysis method was applied for improving existing problem in prediction model and treatment method of basic data. Improving self-regressive mathematical model was established and taken the model identification. Using SPSS software simulates and further assists model identification and using given wind farm historical output power data to forecast one and multi-wind power unit output power in odd-number days and a week. Finally, this paper compares and analyses the getting prediction power and expound the next step work that improves the wind power prediction accuracy.