The Model of Ongoing Diagnosis of Process Faults and Detection of Cybernetic Attacks for a Steam Line

M. Syferta, P. Wnukb, A. Sztyber-Betleya, M. Pobochab aFaculty of Mechatronics, Institute of Automatic Control and Robotics, Warsaw University of Technology, św. A. Boboli 8, 02-525 Warsaw, Poland bEnerbit Sp z o.o., ul. Czerniakowska 26b, 00-714 Warsaw, Poland

Full Text PDF

The paper presents a model of a system for ongoing monitoring and diagnosis of process faults and detection of cybernetic attacks for a section of the boiler steam line. The model consists of the project of an algorithm and a prototype implementation of the system. The implemented algorithm integrates, in one coherent approach, the tasks of monitoring process faults based on parametric partial models and detecting cybernetic attacks based on dedicated checks for control system loops based on a selected set of performance indices. The important model feature is the use of data on malfunctions and detected intrusions obtained from the industrial control system and intrusion detection system during the reconciliation stage and final diagnosis formulation. The operation of the system is presented in the example of a pilot application for a simulator of a section of a steam line. Based on the presented summary of the test results it can be stated that the proposed solution allows for obtaining high detectability and isolability of process faults and their distinguishability from cyber attacks. One of the advantages of the proposed solution is the indirect use of knowledge about phenomena occurring in the process in the task of detecting cybernetic attacks. The structure of the proposed model allows for its relatively easy adaptation to other technological processes.

DOI:10.12693/APhysPolA.146.438 topics: on-line diagnostics, process diagnostics, cybernetic attacks, steam line