Data-driven Analysis and Optimization of Combined Cycle Power Plants using Machine Learning Models

The current global energy demand relies more on Combined Cycle Power Plants (CCPPs) for their high efficiency and reduced environmental footprint. However, the performance of these plants is very sensitive to several environment parameters including temperature, pressure, humidity, and exhaust vacuum. This paper is intended to use machine learning (ML) approach to model and optimize CCPP energy production based on these factors. The proposed method uses a dataset with hourly environmental measurements, to provide detailed analysis using ML techniques including Random Forests and Neural Networks to identify any potential nonlinear relationships and predict energy output. The results showed that ambient temperature has the most significant influence on energy production, followed by vacuum, pressure, and humidity. In addition, this paper also highlighted optimal environmental conditions that maximize energy output, which can help and support power plant operators in optimizing their operation factors. In summary, the recommendations and outcomes of this paper provide necessary steps for integrating advanced ML techniques into CCPP operations, enhancing both efficiency and sustainability.