Monitoring and control of polymer production line based on machine learning

Abstract The work is devoted to the development of an application for monitoring and controlling state equipment (extruder) petrochemical industry based on sensor readings using a machine learning model. statistical relationships technological process parameters are analyzed, most significant influencing occurrence failures determined SHAP values. hypotheses regarding effectiveness various algorithms in relation real problem predicting technical extruder tested. A gradient boosting model has been developed predict probability shutdown due formation polypropylene agglomerates. allows interpreting results model, which makes it possible select important that have greatest impact failure, also proposing prototype system