Model za rano otkrivanje neusaglašenosti procesnih parametara u proizvodnim sistemima

Abstract

Doktorska disertacija se bavi razvojem konceptualnog modela za rano otkrivanje neusaglašenosti procesnih parametara (RONP) u proizvodnim sistemima. RONP model predstavlja hibridni model baziran na upotrebi fazi ekspertnih sistema i metoda napredne analitike, čiji je razvoj podeljen u sedam faza primenom i prilagođavanjem metodologije proučavanja podataka. Verifikacija modela je urađena u procesnoj industriji za proizvodnju podnih obloga od vinila gde je i eksperimentalno potvrđena njegova primenljivost.

The Ph. D. thesis deals with the development of a conceptual model for early detection of non-compliance of process parameters in manufacturing systems. The model represents a hybrid model based on the use of fuzzy expert systems and advanced analytics methods. The development of the model is divided into seven phases by applying and adapting the data minig methodology. The verification of the model was done in the process industry for the production of vinyl flooring, where its applicability was experimentally confirmed.