Study On Multi-scale Power Prediction Model Of BA Russell Wind Farm Based On Combination Algorithm

Enterprise-built wind farms can solve the urgency of enterprise's own load power supply demand,realize complementary advantages and make full use of energy.Prediction of-wind farm power generation can provide a useful basis for the normal operation of wind power grid.At present,power grids in resource-intensive provinces and energy investors in China are actively carrying out relevant research,and the necessary communication basis and meteorological services are all matched.In preparation,however,how to plan and what form to adopt for the wind power forecasting system in China are still being explored.Although forecasting systems in Europe and the United States are becoming m,ore and more mature and can be fully used for reference,the uniqueness of wind power resources can not be copied from the existing systems,but should be closely combined with the characteristics of power grid resources in China to carry out algorithm research.For Balasu wind farm,a wind power prediction and error analysis model is proposed.The main work of this paper is as follows:Firstly,the influencing factors and mechanism of wind power output are analyzed theoretically.Secondly,the general method of wind power forecasting is studied,and the data preprocessing method is put forward The short-term and ultra-short-term multi-time scale wind power forecasting technology is studied and reconstructed,and the multi-combination wind farm power forecasting model is constructed.The domestic numerical weather forecasting data are used to compare the various forecasting models of wind power forecasting,and the error mechanism is analyzed.Finally,according to the actual parameters of the wind farm,a prediction model is established to predict the powder of Balasm wind farm in ultra-short time and short time,and error analysis is carried outThe research results of this paper can provide useful reference for similar enterprises to connect power supply system to wind power generation for power prediction.