ARTIFICIAL NEURAL NETWORK FOR PREDICTION OF THE BEGINNING OF DELAMINATION AT RUBBER CONVEYOR BELTS

1. Biljana Milutinović, College of Applied Technical Sciences Niš, Serbia
2. Petar Djekic, The School of Higher Technical Professional Education, Niš, Serbia
3. Milan Pavlović, College of Applied Technical Sciences Niš, Serbia
4. Mladen Tomić, The School of Higher Technical Professional Education, Niš, Serbia

Indispensable parts of all modern industry systems are conveyor belts. Qualitative and reliable conveyor allows continuous operation of the conveyor system and increases the efficiency of the entire production process. Rubber conveyor belts may contain a steel core or multiple layers of rubber and canvas (carcass). One of the most common causes of failure and main issues in the use and maintenance of rubber conveyor belts with carcass is delamination. In this paper, in application of artificial neural network, the prediction of the beginning of delamination at rubber conveyor belts is presented. A multi-layer feed forward artificial neural network is used and the two input variables considered in the artificial neural network: length of the conveyor belt and the number of layers. Results point that artificial neural network can be trained with exploitation data and later effectively used in predicting of the beginning of delamination at rubber conveyor belts. Thus presented methodology can be used in the control and maintenance of conveyor belts.

Ključne rečitestttt :

Tematska oblast: Maintenance of Engineering Systems and Occupational Safety Engineering

Datum: 27.02.2017.

13th International Conference on Accomplishments in Mechanical and Industrial Engineering

Datoteka uz rad  (2.93 MB)


Ostali radovi sa konferencije


Pretraži radove