Online Scrolling Short Electric Load Forecasting Method Reseach Based On The Measurement Sequence

Electricity load forecasting is one of the important job of the power system,power load forecasting is accurate or not,which is directly related to the development of a regional economy,rational and orderly conduct electricity load forecasting is to ensure that the foundation of a regional development;In recent years,the power load forecasting gradually attracted people’s attention,both from the perspective of the level of energy and urban construction,electric power load forecasting accurate or not have a very big impact on the entire community,and how to improve the accuracy of power load forecasting is very important subject.The study of the power system is very complex,there are differences between different regions,the gap between different industries is also great though in the same area,and how to finance these factors to increase the maximum extent as possible and the power load forecast accuracy is the main problem to be studied.effective solution to this problem is for the rational allocation of the power supply system plays a vital role.How to make the grid system to maximize benefits is the purpose of the study of the power load forecasting.This paper focus on the research of short-term load forecasting values,conducting the research on the power load time series,through the historical data analysis,to establish forecasting electricity load forecasting models for time series,including adaptive AR model based on least squares,scroll adaptive AR model of power load forecasting model based on gray theory and Kalman filtering online.By analyzing the cyclical nature of power load time series,based on the the standard least-squares algorithm,a weighted least squares prediction algorithm is given to redesign the objective function of the system.On the other hand,based on the basis of least squares theory,establishing adaptive AR model.The simulate result proved the online scrolling least squares not only keep real-time,but also improve the accuracy.Grey theory is a commonly used in long-term forecasting methods,this article make the use of short-term day load forecasting.By accumulating data,establish GM(1,1)model to forecast the short-term load time sequence.And a development is given to improve the GM(1,1)model,though accumulating the original data,reestablish the system model.Improving the system forecasting accuracy,and it is suit for small amount of data.On the basis of Kalman filtering theory,the AR model is proposed based on the Kalman filter algorithm and Volterra kernel simulating forecasting model.On the basis,the onlinescroll Volterra kernel simulating forecasting model based on Kalman filtering theory is given to forecast the short-term power load.The online scroll Volterra kernel simulating forecasting model based on Kalman filtering theory is better than the two methods in the real-time and accuracy.Online scrolling shows the real-time of system,and is an important direction for short-term power load forecasting.