CPL - Chalmers Publication Library
| Utbildning | Forskning | Styrkeområden | Om Chalmers | In English In English Ej inloggad.

Load balancing of welds in multi station sheet metal assembly lines

Johan Segeborn (Institutionen för produkt- och produktionsutveckling, Produktutveckling) ; Daniel Segerdahl ; Johan S Carlson ; Anders Carlsson ; Rikard Söderberg (Institutionen för produkt- och produktionsutveckling, Produktutveckling)
Proceedings of the ASME 2010 International Mechanical Engineering Congress & Exposition, Vancouver, British Columbia, Canada, November 12-18, 2010 (2010)
[Konferensbidrag, refereegranskat]

The balancing of weld work load between executing stations and its robots has a significant influence on achievable production rate and equipment utilization. However, no automatic simulation based method for line balancing has been formulated up to this point. In practice, it is still manually conducted. Therefore in this work we propose two novel methods for load balancing of welds in multi station sheet metal assembly lines to minimize line cycle time. The methods are based on superimposition of the scenes/geometries of all line stations, with maintained robot positioning relative to the work piece, creating a “multi station”. The weld load is balanced between all multi station robots, whereupon the individual robots are combined into stations and coordinated station wise for simultaneous operation. Furthermore one of the proposed methods reduces the subsequent need for robot coordination, by introducing some restrictions on the load balancing: Firstly, for each robot, the weld load is balanced over the other station robots such that the working envelopes are maximally separated. Secondly, for each robot, the weld load is balanced over equivalently positioned robots in other line stations, based on previous station load balancing techniques. The proposed line balancing methods are applied on two industrial case studies which each involves the balancing of about 200 automotive stud welds between 3 stations, each of 4 robots. One of the proposed methods produces line cycle times close to that of the slowest uncoordinated robot, which can be considered a theoretical optimum of the line cycle time. Corresponding algorithm running time is about 30 minutes on an Intel Core 2 Quad with 8 GB RAM.

Nyckelord: Sheet metal assembly, equipment utilization, load balancing, robot coordination

Den här publikationen ingår i följande styrkeområden:

Läs mer om Chalmers styrkeområden  

Denna post skapades 2010-11-15. Senast ändrad 2016-05-11.
CPL Pubid: 129119