Improvement of Manufacturing Processes by Artificial Neural Networks Analysis

Abstract

Manufacturing processes consist of activitiesaffected by a large number of variables. Theaim of this study is to show that improvementscan be made by using artificial neural networkmethods at stages of manufacturing such asplanning of processes, forecasting of the futuresituation, monitoring and control. In the study, amanufacturing process with 15 input variables wasmodeled using artificial neural networks, networktraining was provided, and a trained network wasused to obtain the best output performance inthe current situation. Artificial neural networksare useful tools in finding out the consequencesof any change that may occur in variables and inimproving the processes with this way. The resultsshow that artificial neural network models can bewell adapted to manufacturing processes.