Research On Forecasting Method Of Wind Power Installed Capacity Based On Combined Grey Prediction Model

As our country’s "carbon peak" and "carbon neutral" goals are proposed,accelerating the transformation of my country’s energy structure has become an important strategy for the country’s energy development.Wind power is not only clean and renewable,but after years of development,the technology has gradually matured and the cost has been continuously reduced,which plays an important role in promoting the transformation of the energy structure.In October 2020,at the Beijing International Wind Energy Conference,the participants jointly issued the "Wind Energy Beijing Declaration".The "Beijing Declaration on Wind Energy" proposed that,in order to adapt to the national strategy of carbon neutrality,during the "14th Five-Year Plan" period,the annual average installed capacity of wind power will be guaranteed to exceed 50 million kilowatts.To judge that this goal can be achieved,it is necessary to accurately predict the installed capacity of wind power.Doing a good job in wind power development requires scientific planning by the government and scientific demonstration of the project by investors,all of which rely on accurate forecasts of installed wind power capacity.However,the development history of wind power generation in my country is relatively short and presents a leapfrog development history.This has caused the historical data of my country’s wind power installed capacity to show a small sample and non-linear data characteristics,which increases the difficulty of forecasting wind power installed capacity.In addition,the installed capacity of wind power is also affected by power demand,grid construction level and other factors,so these influencing factors should also be considered in the construction of the forecast model.The small sample,non-linear,and multi-factor data characteristics of wind power installed capacity make it difficult for many models to achieve high prediction accuracy.In order to solve this problem,this paper constructs the PSO-FANGBM(1,1)-PLS model,and launches the following research work for the prediction method:First of all,in view of the basic theory and practical difficulties of wind power installed capacity forecasting,this paper constructs the PSO-FANGBM(1,1)-PLS model.Among them,in view of the small sample and non-linear data characteristics of wind power installed capacity historical data,this paper uses the nonlinear gray Bernoulli forecasting model to predict;in order to reduce the volatility of the original data,this paper uses the fractional accumulation operator for data preprocessing;In order to optimize the parameters of the model,particle swarm optimization algorithm(PSO)is introduced into the model construction;at the same time,partial least square regression(PLS)is introduced to add influencing factors to the prediction of the model;And construct the PSO-FANGBM(1,1)-PLS model by combining weighting.After that,in order to verify the accuracy of the model,this paper takes the actual data of my country’s wind power installed capacity and related influencing factors as the research object,and conducts a precision test and comparative analysis of the model.The results show that the PSO-FANGBM(1,1)-PLS model has good prediction accuracy in the prediction of wind power installed capacity,and the prediction accuracy is better than other comparison models.Finally,this paper applies the constructed model.By setting two different development scenarios,we predict and analyze my country’s wind power installed capacity from 2020 to 2025,and combine the goals and requirements of wind power development to give the development of wind power.Countermeasures and suggestions for power generation.