Coupling a chaotically encoded firefly algorithm with ranking to a physics-based mathematical model for robust optimisation of a gas turbine energy system

Abstract

The aim of this study is to probe the potentials of a well-known metaheuristic approach called firefly algorithm with ranking (FAR) for optimising the operating parameters of a complex gas turbine energy system. FAR is a modified version of classic firefly algorithm (FA) which is suited for handling complex constraint optimisation problems. Firstly, by using the first law of thermodynamics, a mathematical model is implemented to analyse the most important design parameters affecting the efficiency of the gas turbine energy system. Thereafter, two well-known chaotic maps, i.e., Gauss and sinusoidal maps, are embedded into the algorithmic structure of FAR to prepare a powerful tool for the considered problem. To ascertain the veracity and the efficacy of the proposed chaos-enhanced FAR (CFAR), a number of chaos-enhanced rival modern optimisers, i.e., chaotic artificial bee colony (CABC), chaotic particle swarm optimisation (CPSO), and chaotic genetic algorithm (CGA), are applied to the considered optimisation problem. The results indicate that CFAR can easily outperform the rival techniques, and yield robust and accurate results.