Research of Artificial Neural Network Based on Gray Relevant Close Degree in the Medium Long-Term Burden Forcasting
The influence factor of the electric power burden forecast is numerous and uncertain, an in recent methods, each of methods can be much too difficult to get higher precision demand. At the relation research of GDP, we find that it has a linear relation with electric power load, so first format data pretreatment model, and then take advantage of GM (1,1), set up grey data pretreatment model. Sebusequently revise data recerved from the data pretreatment models based on improved gray relevant close degree. The revised data are the importation of the neural network model to set up model and train. While forecasting, first forecasting next year electricity consumption, after that take advantage of it to predict the second year s. Finally the instance testifies the usefulness of the method which is applicable to predict to the medium long-term burden and has higher forecasting precision and certainly practical value.
Artificial Neural Network GM (1,1) Gray Relevant Close Degree Linear Regression
Niu Dongxiao Li Yanchang
School of Business and Administration, North China Electric Power University, Baoding P.R.China, 071003
国际会议
天津
英文
2007-10-20（万方平台首次上网日期，不代表论文的发表时间）