Hybrid swarm optimization for vapor–liquid equilibrium modeling

Abstract

A hybrid algorithm based on particle swarm optimization and ant colony optimization was used to describe the vapor–liquid equilibrium of complex mixtures. The proposed PSO+ACO algorithm is tested on several benchmark functions from the usual literature. Firstly, nine binary vapor–liquid phase systems containing supercritical fluids and ionic liquids were evaluated for optimizing the equation of state method. Next, twenty binary vapor–liquid phase systems were described using two activity coefficient models optimized by the hybrid algorithm. The results of vapor–liquid equilibrium modeling were compared with the Levenberg–Marquardt algorithm, and show that the application of PSO+ACO algorithm on thermodynamic models such as equation of state methods and activity coefficient models, is crucial, and that the hybrid PSO+ACO algorithm is a good tool to optimize the interaction parameters to describe the vapor–liquid equilibrium of several systems.