Poincaré plot-based fault detection on tennessee eastman process using various machine learning algorithms Çeşitli makine öğrenimi algoritmalarını kullanarak tennessee eastman sürecinde poincaré grafik tabanlı hata tespiti

Çancıoğlu, Emre

ABSTRACTThis thesis presents diagnose and detect faults of Tennessee Eastman Process (TEP) with machine learning algorithms via Poincaré plots-based feature extraction and statistically analysis-based feature selection. The IEEEDataPort online dataset, obtained from a big plant that contains nonlinear processes from various chemical units, is utilized in this thesis. It contains measures from 52 process points in TEP with 20 dissimilar malfunction types. In this study, raw dataset and dataset that applied feature extraction and feature selection was used. Poincaré plot was applied to the datas ...Daha fazlası