Hourly Temperature Forecasting based on Euclidean Distance Algorithm with Solar Terms

In this paper, a simple and accurate algorithm is proposed for short-term temperature forecasting in power utilities and industrial applications. Solar terms that have been used in Asian countries for centuries are advantageous for the classification of annual historical data. In this paper, an hourly temperature forecasting algorithm based on the Euclidean distance algorithm with 24 solar terms is proposed. In the proposed method, the historical data are classified by the 24 solar terms. The Euclidean distance is used to determine the historical data with most similarity to the historical data in the database. The artificial neural network algorithm was used as a benchmark for comparison purposes. The results show that the proposed scheme has the advantages of simplicity, fast operation, and excellent forecasting performance.