An improved class-shape transformation-based particle swarm optimization (ICST-PSO) algorithm and its application on airfoil aerodynamic performance optimization

Abstract

In order to quickly and accurately obtain and optimize the aerodynamic performance of airfoils, an improved class-shape transformation-based particle swarm optimization (ICST-PSO) algorithm is proposed in this work. The proposed ICST-PSO algorithm is not only used to generate the geometric coordinates of airfoils, but also for optimizing their aerodynamic parameters. Unlike traditional PSO approaches that optimize all CST parameters simultaneously, the proposed method first identifies the four most influential parameters through a systematic sensitivity analysis, and then focuses the optimization exclusively on these key parameters. This strategy effectively reduces the optimization dimensionality while preventing the generation of geometrically invalid airfoils. Both CFD and XFOIL simulation results show that the proposed ICST-PSO algorithm significantly improves the aerodynamic performance of the NACA4415 airfoil, including its lift, drag, flow velocity, pressure distribution, and lift-to-drag ratio at different angles of attack. Specifically, the NACA4415 lift-to-drag ratio ( Cl / Cd ) growth rate of the traditional particle swarm optimization (PSO) method is only 11.6%, while the NACA4415 Cl / Cd growth rate of the proposed ICST-PSO method achieves 17.9%. Moreover, the calculation time of the proposed ICST-PSO method is about 6.3 minutes, compared to 10.2 minutes for traditional PSO, representing a 38.2% improvement in time efficiency.