# Net Hourly Electrical Power Output Prediction in a Combined Cycle Power Plant (CCPP)

## Abstract (300 words maximum)

The design and operation of thermodynamic power plants require a thorough understanding of the complex thermodynamic processes involved. Mathematical models have traditionally been used to analyze such systems, but these models often require a large number of assumptions and parameters to accurately capture the unpredictability of the actual system. As a result, the accuracy of the model may be limited, and the analysis of the power plant may not be as precise as desired. Machine learning and deep learning approaches have recently emerged as promising alternatives to traditional modelling techniques. In contrast to mathematical modeling, machine learning algorithms can learn patterns and relationships from data, without requiring a pre-specified model structure or set of assumptions. This can allow for more accurate and flexible predictions of system behaviour. This project aims to develop an intelligent and accurate electricity prediction system using regression techniques. The objective is to build a deep neural network architecture that can accurately predict the electrical power output generated by the power plant, given a set of input variables. The model would be trained on a large dataset of combined cycle power plant which undergoes exploratory data analysis and data cleaning techniques prior to training. The performance of the developed system will be evaluated using metrics such as mean absolute error, mean squared error, and R-squared. By leveraging the power of deep learning, this system can accurately predict electricity production based on a range of input variables. This project has a potential to not only improve the efficiency of existing power plants but also aid in the development of new, more efficient ones. Ultimately, this project could significantly impact the energy industry and contribute to a more sustainable future.

## Academic department under which the project should be listed

CCSE - Computer Science

## Primary Investigator (PI) Name

Md Abdullah Al Hafiz Khan

Net Hourly Electrical Power Output Prediction in a Combined Cycle Power Plant (CCPP)

The design and operation of thermodynamic power plants require a thorough understanding of the complex thermodynamic processes involved. Mathematical models have traditionally been used to analyze such systems, but these models often require a large number of assumptions and parameters to accurately capture the unpredictability of the actual system. As a result, the accuracy of the model may be limited, and the analysis of the power plant may not be as precise as desired. Machine learning and deep learning approaches have recently emerged as promising alternatives to traditional modelling techniques. In contrast to mathematical modeling, machine learning algorithms can learn patterns and relationships from data, without requiring a pre-specified model structure or set of assumptions. This can allow for more accurate and flexible predictions of system behaviour. This project aims to develop an intelligent and accurate electricity prediction system using regression techniques. The objective is to build a deep neural network architecture that can accurately predict the electrical power output generated by the power plant, given a set of input variables. The model would be trained on a large dataset of combined cycle power plant which undergoes exploratory data analysis and data cleaning techniques prior to training. The performance of the developed system will be evaluated using metrics such as mean absolute error, mean squared error, and R-squared. By leveraging the power of deep learning, this system can accurately predict electricity production based on a range of input variables. This project has a potential to not only improve the efficiency of existing power plants but also aid in the development of new, more efficient ones. Ultimately, this project could significantly impact the energy industry and contribute to a more sustainable future.