Power Load Forecasting Model Based on Harmonic Clustering and Classification Method

Clustering and classification are two important research areas of data mining, and classification needs prior-knowledge, while clustering needs a similar measure to find its inherent characteristics from the data. In real application, the results of classification and clustering are often inconsistent. This paper defined harmonic matrix to solve this problem, and proposed a harmonic clustering-classification algorithm to make the results of classification and clustering keep high consistency. This method has been used in power system load forecasting, and the classification results obtained are more reliable.