Diagnostica Avanzata di Sistemi di Controllo di Processi Industriali - Advanced Diagnosis of Industrial Process Control Systems

This PhD work aims to develop an advanced monitoring and diagnostic system for industrial process control loops. This study is based on techniques and algorithms which make use of data available in industrial plants in order to evaluate performance of control loops, detect and distinguish different causes of malfunction, and suggest counteractions to perform. The overall activity includes modeling and simulation in MATLAB, experimentations on pilot plants, analysis of industrial data, and implementations in process plants. The whole PhD activity is framed in a series of projects for development and analysis of monitoring systems, carried out in the last 15 years within the Laboratory of Control of Chemical Processes (CPCLab). The software developed in this thesis is an evolution of the monitoring system, called PCU (Plant Check Up), which has formed a subject of research in the past several years. Different versions of this monitoring system are now available, which vary depending on equipments and devices used in the plants and on variables and measurements available from DCS. Different collaborations and partnerships with Italian industrial companies (as ENI and ENEL) have been established in the last years. Industrial implementations have constituted an interesting source of inspiration on real problems and a mean of validation of the actual operating ability of the system, as well as a large data base valuable to test new monitoring and diagnostic techniques. These activities have been accompanied by technical reports, and have also generated some scientific papers for the most significant results and applications.