Fault detection in industrial automation systems with artificial intelligence methods

In this thesis; the data sets collected from real production lines are evaluated by using artificial intelligence methods to determine the faults in automation systems.These real datasets collected from production lines are obtained in line with the concepts of "Internet of Things" and "Big Data" which entered our lives with Industry 4.0. These data processed by using Perceptron algorithm (which is an algorithm of Artificial Neural Network), Random Forest algorithm and Gradient Boosting (GBM) algorithms and then the system failures are classified. By analyzing the performances of the artificial intelligence methods, the most appropriate classification method is determined for the selected model. The results of this thesis show that artificial intelligence methods can be used successfully in fault detection for industrial automation systems.