A model-based machine learning to develop a PLC control system for Rumaila degassing stations

Abstract

Degassing station breakdowns can be dangerous to the operator health and the environment.Programmable logic controllers (PLCs) are key modules of manufacturing control systemsthat are applied in the complex oil and gas units to reduce manpower and unnecessary faults.However, feeding a PLC with data is a difficult part due to the need of system log files whichrecords all events that occur in the oil fields and provide visibility to a given environment.Moreover, most critical chemical processing plants and oil distributions are visualized andinspected by Supervisory Control and Data Acquisition Systems (SCADA). These systemshave been focused on safety, and there are issues that they could be the target of worldwideterrorists. Along with the frequently rising internet-related attacks, there is indication that ourdegassing stations may similarly be susceptible; for that reason, it is essential to secure PLCand SCADA from undesired incidents. Recently, machine learning (ML) has been increasinginterest in industrial systems to detect, identify, and store information. Therefore, we proposeto apply an advance ML based on deep neural networks to the PLC system with the purposeof: 1) detecting anomalous or irregular PLC actions; 2) Optimizing the operation of systemsand its facilities; 3) allowing the equipment to respond to changing and novel scenarios; 4)Making predictive maintenance possible. The SIMATIC S7-1214 CPU universal TIAplatform was used as the main decision-making module. Experimental results demonstratethe effectiveness and utility of the proposed approach to process large amounts of dataanalytics and sensor measurements, allows it to spot potential problems and provide possiblesolutions