Analysis of Industrial Network Parameters Using Neural Network Processing
R. F. Gibadullin, D. V. Lekomtsev, M.Yu. Perukhin
Abstract
We used artificial neural networks and diagnostic network information to assess the condition of PROFINET (Process Field Network). An artificial neural network determines whether the network works well. An important part of this work is data preprocessing. This is done using quantization, data alignment, reducing the number of inputs, and other preprocessing techniques to create a new version of the dataset to improve accuracy. The obtained data makes it possible to perform a number of experiments and to find out what approach to data preprocessing shows the best results. The results were evaluated based on two datasets. The first dataset contains diagnostic data of a well-functioning network, and the second one consists of data in which network problems were detected. The highest accuracy obtained in this work is a 98.91% rate of recognizing problems in the network and an accuracy of 87.70% when the network is working well. The work also opens opportunities to improve accuracy in the future.