| dc.description.abstract | Scheduling is a major problem in the job shop world. When scheduling is out of control,
the production schedule is in a constant state of change, and chaos reigns on the shop floor.
Increasing business performance by improving production scheduling is a rapidly growing
area of interest and one that typically generates very substantial business benefits. There is
therefore considerable interest in improving production scheduling to enhance business
performance.
The complexity of production scheduling present a formidable decision task which
traditional approaches to scheduling are inadequate in modeling the complexity of realistic
manufacturing floors, and in handling the diverse set of constraints, the production
objectives, and shop-floor uncertainties.
This research has examined an adaptive scheme for the automated leaning of scheduling
strategies. The learning component is Genetic algorithm based. Genetic algorithms (GA)
mimic the evolution and improvement of life through reproduction.
The results from this research show that using GA to solve the job-shop scheduling is
suitable in this complex field where there are potentially many uncontrollable or
unmeasurable parameters.
The GA model created is capable of generating results which gives an accurate
representation of the real life situation, and offers solutions to help set up efficient new
factories or improve existing inefficient ones. | en |
