Identifying working day and rest day data based on machine learning method for more accurate transformer load forecasting

摘要

We present in this paper an algorithm for identifying the working day load and the rest day load. The proposed algorithm based on Support Vector Machines (SVM) can be used to characterize working day and rest day load. It combines the six feature quantities that can classify the two types of loads very well and has high precision. The application of the trained model and other load data can also be well recognized, which has good versatility and is of great significance in the classification prediction of short-term load forecasting.