Optimal grey topological predicting approach to short-term load forecasting in power system

A refined grey topological predicting approach is proposed in this paper to solve short-term load forecasting (STLF) problems in power systems. The traditional topological predicting technique is not capable of accurately predicting the graphical outline in cases of periodic fluctuation. The proposed method can overcome this disadvantage. The proposed method takes an hour as the threshold value, arranges the power load and temperature data into original data sequences and variable sequences respectively, and establishes a grey GM(1,2) model for load forecasting. In order to construct an optimal GM(1,2) model to raise the accuracy of predicting, the real-code genetic algorithm (GA) is applied to search the optimal value of GM(1,2) model. The daily load of the Taipower system was used to test the proposed GA-GM(1,2) model. The test results showed that the proposed method had better accuracy and practicality than traditional statistics technique.