Application Research On Differential Grey Model In Medium And Long Term Power Load Forecasting

With the development of economy and the continuous increase of power demand,as an important part of the power system,power load forecasting plays an important role.Accurate power load forecast can not only meet people's daily power needs,but also ensure the development of the economy.It also reduces the cost of generating electricity.Power load forecasting is mainly divided into ultra-short-term load forecasting,short-term load forecasting,medium and long-term power load forecasting.For power grid planning,it is mainly medium and long-term power load forecasting that provide the basis for it.Accurate medium and long-term power load forecasting can provide a strong support for power system planning and construction,and plays a decisive role in China's national economic development and power system planning,but the current forecasting accuracy of medium and long-term power load needs to be improved.Because the historical data of medium and long-term electric load forecasting is less,and the gray theory based on "small sample" system is used for medium and long-term electric load forecasting,it has stability and high prediction accuracy.In this paper,the application of multivariate grey model optimized by differential evolution algorithm in medium and long-term electric load forecasting is taken as the main research content.Firstly,aiming at the problem that the regularity of historical data is not strong,the variable weight buffer operator is used to preprocess the data.At the same time,the qualitative analysis of the system can be incorporated into the prediction model by using the variable weight buffer operator.Secondly,considering that the power load has many influencing factors,this paper uses a multivariate grey model.The background value is obtained for the multivariate gray model by computing the average value of the two adjacent numbers of the first-order accumulation sequence can be used to obtain a large error.This paper introduces a parameter to determine the proportion of two values when obtaining the background value.The generated parameters are obtained by using a differential evolution algorithm with a minimum of the average of the relative error.What's more,in view of the fact that the electricity consumption in the new year is more closely related to the load in the most recent year,this paper uses the isodimensional idea to predict the electric load.Finally,the optimized multivariate grey model is used in MATLAB for simulation experiments using the relevant historical data of jiangsu province,and compares with the prediction results of GM(1,1)model and MGM(1,m)model,the experimental results show that the method adopted in this paper is obviously better than the other two methods.Based on the above research content,the algorithm is applied in practice and I design a medium and long-term power load forecasting system.First analyzing the requirements of the system,then designing the logical structure,database and function module according to the requirements,and finally developing the system in myeclipse10.After testing,the developed system meets the requirements.