Improved TLBO and JAYA algorithms to solve new fuzzy flexible job-shop scheduling problems

Flexible job-shop scheduling problem (FJSP) finds significant interest in the field of scheduling in dealing with complexity, solution methodology and, industrial applications. However, most of the studies on FJSP, consider the processing time of operations to be deterministic and known at priori while solving the problem. Since uncertainty is bound to occur in industries, deterministic approaches for solving FJSP may not yield good solutions. Schedules generated considering uncertainties may help the manufacturing firms to handle the uncertainties efficiently. The present work aims at solving FJSP in a realistic manner, considering uncertainty in the processing times. A modified version of optimization algorithms without tuning parameters such as teaching-learning-based optimization (TLBO) and JAYA is proposed to solve fuzzy FJSP (FFJSP) with less computational burden. Although there are enough challenging benchmark problems for deterministic FJSP problems, only limited benchmarks are available for a fuzzy variant of FJSP. The currently available FFJSP problems in the literature are small in size as compared to Brandimate data instances which are widely accepted for a deterministic variant of FJSP. Therefore, an attempt has been made in this paper to solve the instances of Kacem’s and Brandimarte’s by converting them into fuzzy FJSP. The present work also provides new challenging problems compared to the existing benchmark problems to study FFJSP.