The facility layout problem (FL) is a non-linear, NP-complete problem whose complexity is derived from the vast solution space generated by multiple variables and interdependent factors. For reconfigurable, agile facilities the problem is compounded by parallelism (simultaneity of operations) and scheduling issues. Previous work has either concentrated on conventional (linear or branched) facility layout design, or has not considered the issues of agile, reconfigurable facilities and scheduling. This work is the first comprehensive methodology incorporating the design and scheduling of parallel cellular facilities for the purpose of easy and rapid reconfiguration in the increasingly demanding world of agile manufacturing. A novel three-stage algorithm is described for the design of acyclic (asynchronous), bufferless, parallel, multi-process and mixed-model production facilities for spaceframe-based vehicles. Data input begins with vehicle part processing and volume requirements from multiple models and includes time, budget and space constraints. The algorithm consists of a powerful combination of a guided cell formation stage, iterative solution improvement searches and design stage scheduling. The improvement iterations utilise a modified (rules-based) Tabu search applied to a constant-flow group technology, while the design stage scheduling is done by the use of genetic algorithms. The objective-based solution optimisation direction is not random but guided, based on measurement criteria from simulation. The end product is the selection and graphic presentation of the best solution out of a database of feasible ones. The case is presented in the form of an executable program and three real world industrial examples are included. The results provide evidence that good solutions can be found to this new type and size of heavily constrained problem within a reasonable amount of time.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:524630 |
| Date | January 2001 |
| Creators | Tebaldi, Enrico |
| Publisher | University of Warwick |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://wrap.warwick.ac.uk/3069/ |
Page generated in 0.0015 seconds