Application of hybrid artificial bee colony algorithm based on load balancing in aerospace composite material manufacturing. (1st April 2023)

• Record Type: Journal Article
• Authors: Wang, Yufang; Ge, Jiarong; Miao, Sheng; Jiang, Tianhua; Shen, Xiaoning
• Is Part Of: Expert systems with applications. Volume 215(2023)
• Publication Date: 2023-04-01
• Subjects: Flexible job-shop scheduling -- Artificial bee colony algorithm -- Aerospace composite manufacturing -- Heuristic rules -- Load balancing -- Multi-objective optimization; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2022.119375
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: A HMABC algorithm based on load balance was proposed. The scouter bee mechanism solves the problem of falling into local optimum. The THR based on machine load balance improves local search ability. The HMABC is applied successfully into aviation composite materials manufacturing. Abstract: The manufacturing of composite materials for the aerospace industry represents a cornerstone of future development in this advanced field. In this study, a mathematical model of flexible job-shop scheduling problem (FJSP) was established by extracting the key manufacturing information from an aerospace composite manufacturing workshop. Intelligent algorithms can effectively solve FJSP. However, when applied to a multi-objective FJSP (MOFJSP), they suffer from drawbacks such as blind search and low efficiency. Therefore, this study proposes a new artificial bee colony algorithm and implements it as part of a hybrid method (referred to as the HMABC) to solve MOFJSP. The method involves three steps: (1) A multiple-rule initial population is constructed to minimize randomness and improve quality. (2) A greedy decoding method is used to decode each individual to further narrow the search range to replace the conventional decoding method. (3) By considering the machine load index (which has often been neglected), the concept of machine load rate is proposed, and three heuristic strategies are designed (with the influences ranging from small to high) to diverge the solution in all directions instead of just a direction. The performance of the proposed method is compared with those of existing algorithms using a classical FJSP test example, and the effectiveness and efficiency of HMABC is verified. Finally, the method is applied to an actual aerospace composite manufacturing system to realize the scheduling of actual production activities.