Application of metaheuristic algorithms in nano-process parameter optimization

This paper presents the adaptation of Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Gravitational Search Algorithm (GSA) in solving a nano-process parameter optimization problem. The nano-process in this study is involving the RF magnetron sputtering process. The performances of the algorithms are compared in this optimization problem. The performance of GA, PSO and GSA is evaluated based on the fitness of the optimized parameter combination, processing times and the results from comparison with the actual laboratory experiments. The purpose of this computational experiment is to obtain the most optimized parameter combination among the selected datasets. The source material used in this study is zinc oxide (ZnO) and the most optimized combination of the process parameters is expected to produce the desirable nanostructured ZnO thin film's electrical properties. The results have shown that GA could perform better than PSO and GSA by generating higher fitness values in 30 trial runs. However, GA has obtained the slowest execution time among the three algorithms. In this study, GSA has also produced an acceptable and promising result with faster execution time. When compared with the actual laboratory experiment, GA and GSA have generated more accurate optimization results. In terms of convergence of the algorithms, GA and GSA have shown more stable convergence compared to PSO. This study has shown that metaheuristic techniques are promising and reliable to be applied in solving this process parameter optimization problem.