Study On The Prediction Of Electric Power Consumption In Power Marketing Systems In Ying Kou Area

Normal and stable operation of power system is one of the most important guarantee of the sustainable development of national economy. With the deepening reform of the power system, power system marketization pace becomes more and more quickly. Power industry has entered the rapid development stage. The power market in China changes from sell leading to purchase leading. In this trend, the traditional management system and management methods cannot meet the need. It is necessary to apply the efficient power marketing decision support system. Prediction of power demand is the base of correct decision and is an important component of decision support system. It has the crucial effect on the electric power marketing decision making. Based on the characteristics of the fuzzy algorithm, fuzzy linear regression forecast model can reduce the dependence on the accuracy of history data and it has been applied in the power system electric consumption prediction widely. However, the prediction accuracy of linear regression model depends on the ambiguity, which relies on the real data. When the data is more irregular, the ambiguity is higher and the prediction precision is lower. In practice, many factors cause the data of the irregular, such as misreading, accidental phenomena and so on. What is more, the irregular data is inevitable. In order to solve this problem, this paper based on the genetic algorithm, proposed a new kind of fuzzy linear regression forecast model. Firstly deleting the irregular data by the genetic algorithm and then build the mathematical model. After simulation the electric consumption in YingKou area in the last ten years. We can get that the new model has lower ambiguity and correspondingly has higher prediction precision. So it is able to improve the performance of predicting electric consumption and promote the economic benefits and social benefits of electric power company.