Comparing Genetic Algorithms and Ant Colony in Optimizing Production Output Function in Dynamic Manufacturing Systems

Abstract

Advance of technology and the increased quality of people's lives through industrial achievements has created an intense competition between manufacturers in providing high-quality products and services. Thus, if a manufacturing company cannot keep the quality and capability of its products up to certain standards, it will eventually have to leave this cycle of production and competition. Usually, reputable manufacturing companies require different techniques to enhance the reliability of their products in this competition. One of the most important elements of industrial products is the product's performance over time, in a way that it can perform its function properly. Considering that system performance over time is a random phenomenon, the role of probability and statistics in analyzing random characteristics of system performance is highlighted. Employing manufacturing systems without paying attention to their reliability increases the chance of sudden failures. Therefore, in this research and by considering time limits, we analyzed production in dynamic manufacturing systems with the aim of minimizing system costs, product's transfer time and pollutants. Sensitivity Analysis (SA) was carried out on Matlab for three different factories with small, medium and large sizes using metaheuristic genetic and ant colony algorithms. Results indicate that genetic algorithm is better than ant colony; because it leads to better, easier and faster optimization of time and cost in this research.