Particle Swarm Optimization Algorithm Parameter Tuning Using Taguchi Method for Gliding Trajectory Optimization of Missile

Abstract

The proposed research approach primarily covers the optimization of particle swarm optimization (PSO) algorithm control parameters for gliding trajectory optimization using Taguchi method. The PSO parameters such as population size, inertia weight and acceleration coefficients are selected for this study. The experiments have been designed as per Taguchi’s design of experiments using L25 orthogonal array. Methodical reasoning ability of Taguchi design is utilized to obtain better population size, inertia weight and acceleration coefficients and consequently, enhance the performance of PSO algorithm for gliding trajectory optimization. Gliding trajectory is optimized by discretizing the control parameter (angle of attack), followed by conversion of optimal control problem to nonlinear programming problem (NLP) and finally solving the problem to realize maximum gliding range. PSO is utilized to solve NLP problem to achieve optimal angle of attack. Analysis of variance (ANOVA) is carried out to estimate significance factors on trajectory optimization. Acceleration coefficients observed to be influencing parameter in gliding trajectory optimization.  Artificial neural network (ANN) method was also exploited to evaluate the significance of various PSO parameters in achieving optimized results. Simulation results portrayed that the gliding range is maximized after PSO parameter tuning. It is observed that the gliding distance of missile improved compared to earlier one. Additionally, simulation results portray the efficiency of proposed method via different test scenarios.

Keywords: Gliding trajectory, Particle swarm optimization, Taguchi method, ANOVA, artificial neural network