Predicting the Energy Efficiency of Thermal Power Plant

with the advent of advance AI based Algorithms, decision making for energy planning in power sector has become simplified due to predicting of energy efficiency of power plants in developing countries. This paper presents a method of predicting energy efficiency of Jamshoro thermal power plant based on historical data of power generating units. There are three steps involved in the process of efficiency of plant. In the first step, preprocessing of raw data gathered from power generating units, auxiliary consumption, net electric output of system and related components of power plant is involved through data mining techniques. The first step screen out useful data and remove redundancy. In the second step preprocessed data is feed into prediction algorithms for learning from data in training phase for model building which involves various predicting algorithms. In the last step, the performance of each model is analyzed, and best fit model with high performance & maximum accuracy with respect to the measured data is suggested for future energy efficiency prediction of power plant. Test results & comparison of each prediction model presented in the end of this paper.