Use Data Mining Techniques to Identify Parameters That Influence Generated Power in Thermal Power Plant

Abstract

The goal of this paper is to identify the parameters that influence the amount of power generated by steam power plants. Data mining tools were used to prove that influencing parameters are differ according to the current status of power plant. Waikato environment for Knowledge analysis (Weka) was used for feature selection and building the prediction model. An initial comparison between many algorithms for each data set was reported. Then the prediction model was built using linear regression algorithm, because it shows the highest correlation coefficient between parameters, and minimum errors. The selected model predicts the generated power using all available parameters as predictors. Although this is not a practical method for power prediction, because not all predictors are controllable, but it reflects how much a parameter influence the amount of generated power. Evaluation results of these models were discussed and a detailed analysis sheet was prepared, to prove that data mining is the best way to predict the amount of generated power, and show the health status of steam power plants.