A MODIFIED HEURISTIC JOB SHOP SCHEDULİNG ALGORUITHM İN AUTOMATED MANUFACTURİNG SYSTEMS

The schcduling problem has been a ehailenge to researehers and manufacturers for several decades. With advanccs in tcchnology, thc associatcd diffîcultics tend to be nıore sophisticated. This creates an inercasing need for improved usagc of thc costly maclıincry. Hcncc Ihc melhodologics for modelling a scheduling problem, i.e., representalion and manipulation of scheduling informalion, gain a nc\v importance (Turkscn et al. 1992). The schcduling funetion plays an importaııl role in automated maııufacturing systems (AMS) and cspccially in Flcxib!c Manufacturing Systems (FMSs), Hovvcver, AMS schcduling is Iremendously compiex duc lo combinatorial explosion, technological conslraints and goals to be achieved (Alptekin and Rabelo 1992). Tiıerc havc been numcrous stud.ics by operation researeh and artifıciaİ intclligencc researehers on thc schcduling problem in manufacturing systems över ten years. Research in scheduling has focused on understanding Ihe variety of schcduling environments that cxist, and construeting scheduling modcls specific to these particular cases. Four typcs of schcduling problem arc distinguishcd in thc litcralurc:singlc machinc-single operation, parallcl machines-single operation, flo\vshop scries of machincs-multiple operations, job shop netvvork of machincs-multiple operations (Turkscn et al. 1992). Job shop scheduling problems in AMS can be solved by follmving thrcc types of melhods: a) exhaustive methods (0-1 algoritlım ete), b) heuristic criteria (list scheduling, shifling bottleneck), c) natural aigorithms (i.e simulated annealing, genetic aigorithms) (Alfano ct al. 1994). To date numerous papers have been pubiished on the job shop schcduling in AMS using heuristic knowledge-based systems. Erschler and Esquirol (1986) presented a job-shop scheduling system, MASCOT, which uses a constraint-based analysis. An expert scheduling system to the preecding was presented in Bensana et al. (1986). The job shop scheduling system, OPAL, integrales the constraintbased analysis modüle with the rule-based decision support modüle. The control strategy of the decision-support modüle is based on the fuzzy set methodology. Subramanyam and Askin (1986) discussed an approach for scheduling an FMS on a daily basis for t\vo shifls to meel the vveckly produetion requircmcnl Sha\v and Winston (1985) studied the planning and control problem in a cellular flexible manufacturing system as a general job shop scheduling. ISIS, deveioped by Fox (1983) at thc Carnegie Mellon University, is a well-kno\vn expert scheduling system for large-scale job shops. KBSS is deveioped and presented by Kusiak (1990) for job shop scheduling in AMS environment (Kusiak 1990). Modifıed algoritlım in this paper is taken KBSS as a 'skeleton' bul instead of using inference engine and individual rules L R A rule and new rule combinalİons are used.