حل مساله زمان بندي كار كارگاهي چند هدفي انعطاف پذير پويا به وسيله الگوريتم ژنتيك توسعه يافته

Solving Multi-Objective Flexible Dynamic fob-Shop Scheduling Problem with Improved Genetic Algorithm

In this paper; Multi-Objective Flexible Job-Shop scheduling with Parallel Machines in Dynamic manufacturing environment (MO-FDJSPM) is investigated. Moreover considering dynamical job-shop environment (jobs arrived in non-zero time), It contains two kinds of flexibility which is effective for improving operational manufacturing systems. The non-flexibility leads to scheduling program which have problems like useless loading machines, bottleneck machines, decreasing desirability sources and a poor function in just in time delivery. Regarding to the flexibility in manufacturing systems, a job could be processed not only in several stations (operational flexibility) but also on several parallel machines in each station (flexibility of parallel machines) which both of them are considered in this paper. In the recent researches about FJSPM and FDJSPM, the single objective models were assessed. Whereas in competitive conditions, decision-makers encountered with simultaneous multi-objective problems that a number of them could be completely conflict with each other. In this research, the objectives are make span, mean flow time and mean tardiness. These objectives are adaptable to the concept of just-in-time and supply chain management. Since the problem is NP-hard, an improved Genetic Algorithm is proposed. Proposed GA compare with Genetic Programming (GP) and GA, the result demonstrate inherence proposed GA. The control parameters in proposed GA are dynamic and changed through the algorithm that leads to reducing the probability of early convergence and local optimum. The mean results for three flexibility levels show that there is 4.9%, 5.33% <£ 4.6% improvement in proposed GA compared with previous results.