Multi-objective optimal control of an industrial absorption column via a hybrid particle swarm optimization–NSGA-II algorithm

Abstract

This paper addresses the dynamic modeling, simulation, and multi-objective optimal control of an industrial packed absorption column for CO₂ removal from natural gas using methyldiethanolamine (MDEA) as a reactive solvent. A comprehensive nonlinear dynamic model is developed to describe the coupled mass and energy transfer phenomena, incorporating axial dispersion, vapor–liquid equilibrium, and temperature gradients along the column. The control objective is to optimally regulate the solvent flow rate and thermal profile in order to maximize CO₂ absorption efficiency while minimizing energy consumption. The problem is formulated under operational constraints, including a first-order dynamic constraint imposed on the outlet CO₂ concentration to ensure smooth transient behavior and avoid abrupt variations that may compromise process stability and product quality. To solve this problem, a hybrid metaheuristic optimization framework combining Particle Swarm Optimization (PSO) and the Non-dominated Sorting Genetic Algorithm II (NSGA-II) is proposed. PSO enhances convergence speed toward promising regions of the search space, whereas NSGA-II ensures a robust exploration and accurate approximation of the Pareto-optimal front. This hybrid strategy enables the simultaneous optimization of conflicting objectives while satisfying the imposed dynamic constraint. Simulation results demonstrate that the proposed PSO–NSGA-II approach achieves improved convergence performance and provides a well-distributed set of Pareto-optimal solutions, offering a balanced trade-off between energy efficiency and CO₂ removal performance. The study highlights the effectiveness of hybrid metaheuristic based optimal control in improving operational efficiency, ensuring process stability, and meeting environmental requirements in industrial gas treatment systems.