Research on the technical framework and critical path of new energy portfolio prediction based on multi-algorithm fusion

In recent years, the new energy power prediction technology has been developing continuously, but there is still the problem of new energy's relatively fragile ability to tolerate extreme weather in actual operation. Therefore, this paper proposes a combination prediction model based on the ordered weighted average operator to improve the accuracy of new energy power prediction under complex weather for the actual operation and production needs of power dispatch. According to the basic process of wind and solar power prediction, the Shapley value method is utilized to calculate the combination prediction weights. The Logistic model, time series ARMA model, and gray prediction GM (1, 1) model are used as the single prediction models constituting the combined prediction model, and the induced ordered weighted average IOWA operator is introduced to establish a new combined prediction model by assigning high and low ranking to the fitting accuracy of the single prediction methods. Aiming at the seasonal and daily characteristics of PV power, the influence of different weather types on the prediction error of PV power is investigated. Comparison of prediction model accuracy and error value is carried out to analyze the prediction effect of the new energy power combination prediction model based on the induced ordered weighted average operator proposed in this paper. The combination prediction model proposed in this paper maintains a high level of prediction accuracy for PV power and wind power under different weather conditions. The overall fluctuation range of its absolute error value in the ultra-short-term prediction of wind power is kept within 0~3. The new energy power combination prediction model based on the fusion of multiple algorithms designed in this paper can improve the accuracy of new energy power prediction in multi-dimensional scenarios, and can provide support for the dispatch operation of new energy-based power systems under complex market environments.