Thermal Power Plants Performance Assessment Optimization using Hybrid Evolutionary Model

Thermal power plant assessment optimization is an environment friendly approach aiming at minimizing the use of thermal power. This study investigates the performance assessment and optimization of thermal power plants in Iran using advanced optimization techniques. The dataset consists of data from 40 thermal power plants collected between 1997 and 2000, covering important input parameters such as installed power, fuel consumption, labor, internal power, forced outage time, and operation time, with total power generation as the output parameter. Min-max normalization of all values were done to ensure consistency and assignment of weights to input parameters based on their importance. Three optimization algorithms were implemented to minimize the DEA BCC objective function: Particle Swarm Optimization (PSO), Firefly algorithm (FA), and PSO-Firefly algorithm (PSO-FA). Each algorithm was evaluated based on the best score obtained in a series of iterations, with the aim of improving its efficiency. The results showed that the PSO-FA Hybrid algorithm outperformed the individual PSO and FA methods in terms of convergence speed and solution quality.