Research Of Grey Prediction Model And Its Application Of Power Load Prediction

Grey System Theory was founded by a Chinese scholar Pro. Deng J.L. It's a new way to study the uncertainty problem of less data and poor information. Grey Prediction Model of it is with the characters of less date, high prediction precision and without prior information. However, more and more studies have shown that GM (1, 1), as the most widely used Grey Prediction Model, is a kind of unbiased exponential model and it couldn't get the expected results sometimes.According to the facts, three new modified methods of Grey Prediction Model are advanced in the paper on the basis of studying the essential conceptions, basic principles and fundamental methods of Grey System Theory. What's more, the specific and practical examples are used to predict in the paper so that to improve its prediction precision and enlarge its practical area, which involve several aspects as follows: 1) eliminated the two limiting conditions of Residual GM(1,1) so that any residual sequence could be disposed by the resembling residual GM(1,1); 2) adopted sequence operator to raise its precision on condition that GM(1, 1) is with low precision when the date sequence increased or decreased too quickly;3)used AGM(1,1) to modify the parameters estimation of GM(1,1) so that to improve the prediction precision of GM(1,1).Besides, a power load prediction simulation system based on Grey System Theory has been designed and accomplished in this paper which takes the Visual Studio.Net 2003 and Access 2003 as development tools. The simulation system can validate the feasibility and validity of the improved prediction model, and can satisfy the actual needs in some concerning departments to provide scientific theory and techniques supports much more conveniently for its prediction works. It applied into a province's Power System to predict the future power load in the province. The results show that: firstly, Grey Prediction Model is of simple process and effective practicality. It could provide scientific theory thoughts for the concerning governments; secondly, the improved one has high precision and less prediction errors; last but not least, it shows that the prediction system of power load based on Grey System Theory is stable and practical, and it could offer some worthwhile references for the work of power load prediction in the Power System in future.