Application of grey Verhulst model in mid-long term load forecasting

Abstract

Mid-long term power load forecasting takes 5 or 10 years as an unit for forecasting. There are several methods for mid-long term load forecasting, in which the gray system prediction method is more effective. GM (1, 1) model is the most commonly used model of gray system. This model which has some limitations in real prediction, does not apply to predict the object has an S-shaped curve of growth. In this paper, the gray Verhulst model is applied to load forecasting, and predict the forecast of actual grid. This paper studies the application of gray Verhulst model in the mid-long term load forecasting: Firstly, it has been proved by examples that the gray Verhulst has higher accuracy than GM (1, 1) model when the forecast has an S-shaped curve. And the result based on the grey Verhulst model can be high precision prediction. Secondly, we applied the gray Verhulst model as a single model to the combination forecasting, and adopted recursive equal weighting method to calculate the weight of each single model that participate in the combination forecasting. Finally, we drew the following conclusion that it is feasible for the grey Verhulst model to involve in the combination forecasting and at the same time when the grey Verhulst model is applied to the combination forecasting which predict the mid-long term load, ideal result can be achieved.