Solving scheduling problem in a distributed manufacturing system using a discrete fruit fly optimization algorithm

This study attempts to optimize the scheduling decision to save production cost (e.g., energy consumption) in a distributed manufacturing environment that comprises multiple distributed factories and where each factory has one flow shop with blocking constraints. A new scheduling optimization model is developed based on a discrete fruit fly optimization algorithm (DFOA). In this new evolutionary optimization method, three heuristic methods were proposed to initialize the DFOA model with good quality and diversity. In the smell‐based search phase of DFOA, four neighborhood structures according to factory reassignment and job sequencing adjustment were designed to help explore a larger solution space. Furthermore, two local search methods were incorporated into the framework of variable neighborhood descent (VND) to enhance exploitation. In the vision‐based search phase, an effective update criterion was developed. Hence, the proposed DFOA has a large probability to find an optimal solution to the scheduling optimization problem. Experimental validation was performed to evaluate the effectiveness of the proposed initialization schemes, neighborhood strategy, and local search methods. Additionally, the proposed DFOA was compared with well‐known heuristics and metaheuristics on small‐scale and large‐scale test instances. The analysis results demonstrate that the search and optimization ability of the proposed DFOA is superior to well‐known algorithms on precision and convergence.