INTEGRATED APPROACH OF SCHEDULING A FLEXIBLE JOB SHOP USING ENHANCED FIREFLY AND HYBRID FLOWER POLLINATION ALGORITHMS

Abstract

Manufacturing industries are undergoing tremendous transformation due to Industry 4.0. Flexibility, consumer demands, product customization, high product quality, and reduced delivery times are mandatory for the survival of a manufacturing plant, for which scheduling plays a major role. A job shop problem modified with flexibility is called flexible job shop scheduling. It is an integral part of smart manufacturing. This study aims to optimize scheduling using an integrated approach, where assigning machines and their routing are concurrently performed. Two hybrid methods have been proposed: 1) The Hybrid Adaptive Firefly Algorithm (HAdFA) and 2) Hybrid Flower Pollination Algorithm (HFPA). To address the premature convergence problem inherent in the classic firefly algorithm, the proposed HAdFA employs two novel adaptive strategies: employing an adaptive randomization parameter (α), which dynamically modifies at each step, and Gray relational analysis updates firefly at each step, thereby maintaining a balance between diversification and intensification. HFPA is inspired by the pollination strategy of flowers. Additionally, both HAdFA and HFPA are incorporated with a local search technique of enhanced simulated annealing to accelerate the algorithm and prevent local optima entrapment. Tests on standard benchmark cases have been performed to demonstrate the proposed algorithm’s efficacy. The proposed HAdFA surpasses the performance of the HFPA and other metaheuristics found in the literature. A case study was conducted to further authenticate the efficiency of our algorithm. Our algorithm significantly improves convergence speed and enables the exploration of a large number of rich optimal solutions.