Master'sOpen Access

Time forecasting with artificial intelligence for products which are manufactured with similar processes

2011
0 views
0 downloads
Advisor: Doç. Dr. Ergün Eraslan

Abstract (EN)

Standart time is the basic requirement in production and capacity plans of thecompanies, although the measurement is very cost-effective, time consuming andrequires a worw including teamwork. In this case, it is necessary today, thedevelopment of alternative methods of measurement.If standart time predicted incorrectly and making future plans with using thisestimation, company can finish work before the estimated and this cause to idlemachines or company can not finish work until the deadline. Company loosemoney in both cases.Recently neural networks are used in every fields where computer and otomationis used and neural networks gives accurate and fast results. Especially, neuralnetworks is used in robotic researches and smart systems and it also is used fortime forecasting with using past data.In this study, an alternative time prediction method is developed for a companywhich collects time with time study. Backpropagation is used during solving thisproblem and an application which lets user to design his own neural networkmodel. In addition application export results and graph desired model results withthe user interface.

Author

Soner Şükrü Altın

How to Cite

Soner Şükrü Altın (Master Thesis). Time forecasting with artificial intelligence for products which are manufactured with similar processes, 2011, Başkent University, Endüstri Mühendisliği Bölümü.

License

Tüm Hakları Saklıdır

This work is shared under the specified license terms.

More theses from Başkent University