Improved Immune Algorithm for Optimizing Distributed Production Scheduling Problem in Flexible Manufacturing System Subject to Machine Maintenance

Competitiveness and rapid expansion of flexible manufacturing system (FMS), as one of the industrial alternatives, has attracted the attention of many practitioners and academicians. Globalization has further encouraged FMS development into distributed, self-reliant units of production center. Flexible manufacturing system in distributed system (FMSDS) considers multi-factory environments, where jobs are processed by a system of FMSs. Scheduling problems in FMSDS deal with allocation of jobs to factories, independent assignment of job operation to machines, and operations sequencing on a machine. In addition, many previous studies neglect the impact of maintenance as one of the core parts of production scheduling. Maintenance significantly affects the overall performance of production scheduling. As such, maintenance is considered in this paper as part of production scheduling. This paper aims to minimize the global makespan over all the factories. This paper proposes an improved immune algorithm (IIA) to solve scheduling problems in FMSDS subjects to machine maintenance. Antibody encoding is adopted to explicitly represents information about a factory, job, and maintenance, while a greedy decoding procedure exploits scheduling flexibility and determines the job routings. Instead of the traditional mutation operator, an improvised mutation operator is used to improve the solutions by refining the most promising individuals within passing generation (iteration). The proposed approach was compared with an ant colony algorithm and several variants of genetic algorithms. The proposed IIA improved the global makespan from 4% to 33% compared with other algorithms. IIA performance has also been tested with several adjustments of population sizes, clonal selection rates, and local explorative mutation rates. Keywords—distributed manufacturing, flexible manufacturing system, artificial immune system, preventive maintenance, optimization