A Spatial Load Forecasting Method Based On Cloud Theory And Analytic Hierarchy Process

This paper presents a spatial load forecasting method based on cloud theory and analytic hierarchy process. Spatial load forecasting is the basis of distribution network planning, and it hinges on choosing suitable load density. Because load density is influenced by many uncertain factors, how to select the suitable load density becomes a complicated decision problem.Based on large amounts of data investigated, this paper selects the influencing factors which play important roles in the load density. Firstly, establish cloud generators and inference rules to quantize the qualitative factors, and then calculate the objective weights of the influencing factors. Secondly, establish hierarchical analysis models, form judgment matrixes according to the experience of experts and calculate the subjective weights of influencing factors. Thirdly, calculate synthetical weights based on objective weights and subjective weights. Finally, using the synthetical weights to forecast the optimal load density and the value of load based on sample data and then take the superimposition curve as the total load curve of the predicted area.The synthetical weights proposed in this paper can take account of the objective regularity and the experience of experts on the premise of considering many influencing factors. The selecting of load density bases on a large number of sample data. This paper expands limited sample data by cloud generators so that overcomes the difficulty of getting sample data. Inputting the values of influencing factors in a given year of the area to be predicted, the model can work out the load density of the area in the given year. This method can not only predict the spatial distribution of the electric power load, but also predict the evolution process of the spatial load.An example demonstrates the superiority of the proposed model in spatial load forecasting.