Printed Circuit Board Assembly Planning for Multi-Head Gantry SMT Machine Using Multi-Swarm and Discrete Firefly Algorithm

Abstract

This research proposes a Multi-swarm and Discrete Firefly Algorithm (MDFA) to deal with the component sequencing problem (CSP) and feeder assignment problem (FAP) simultaneously for a multi-head gantry surface-mounted technology (SMT) machine. The MDFA is an improved version of standard Firefly Algorithm (FA) which is a kind of nature-inspired and population-based metaheuristic. To our best knowledge, the FA has never been used to deal with the CSP and FAP simultaneously for the multi-head gantry SMT machine. Empowered by novel features such as multiple swarms as well as adaptive and discrete moving step, in addition to using an exploration-to-exploitation strategy, the MDFA is found capable of searching a solution space effectively. To investigate its effectiveness, the MDFA has been compared to the standard FA, particle swarm optimization (PSO), and genetic algorithm (GA). The experimental results show that the MDFA outperforms the others in terms of assembly time.