Application Of Grey System Theory In Power Load Forecasting

Electric Load forecasting is an important task of power utilities. It is the important foundation of the study on electric system planning problem, economical running. The accurate prediction of electric load is significant to the national economic development and the safe economical operation of power system. It has become one of the important tasks in the study on modern electric system. At first, this text has briefly introduced meanings, the current development and the main methods which the electric load forecasting are used. This paper aimed at the floatability which the load data had and the nature that neural networks model is easy to fall into part infinitesimal value and low iterative speed, this paper proposed a method based on neural networks and SVM. SVM can dispel the less summation and the larger single point errors of neural networks, and the neural networks could dispel the simplification of the model of the vector support machine. Because the electric load data have exponential and periodic nature, while general Grey models can't be easy to deal with this kind of relation. Provide high-order grey dispersed array models of two steps and three steps. Grey models of two steps can show the periodic law of load data or the exponential law well, Models of three steps can give consideration to two kinds of characteristics of the load data, Thus these models could consider the nature of the load data in an all-round way. This paper has unitized various kinds of influence factors by way of influencing factors. Aimed at the periodic nature of the load data, proposed a load forecasting method that based on temporal sequence and used grey relevancy ideas. So this method could weaken the influence of the grey model that can't consider the periodic nature of the load data. Furthermore, the arithmetic example of forecasting method is analyzed, The result can meet the expectant request.