Mehanna S. Computer systems and nets diagnosis methods by using real-time expert systems

The object of research - the processes of diagnosing faulty functional states of computer systems and networks. The purpose of the research - intelligently diagnose faulty states of computer systems based on expert techniques and fuzzy logic with the use of structural models. Methods of research - to develop a method for troubleshooting and fault finding techniques used graph theory, set theory, technical diagnostics of digital systems and fuzzy logic. The scientific novelty - first developed a hierarchical model of computer systems for the structural level in the form of a multi-level Petri nets, which allows you to create a diagnostic model with faulty states for digital systems of any complexity, the first time developed a method for diagnosing computer systems using the theory of multisets, which includes the steps of the analysis of a limited number of diagnostic parameters and decision-making in the presence of incomplete and contradictory information that allows you to diagnose the faulty state of a computer system with a great performance in real time; improved method for the analysis of production rules of expert systems for diagnosis of functional ability of a computer system, which, unlike the existing uses binary ratio of the characteristic values of the parameters for signs of diagnoses during the synthesis of expert systems, reducing the number of production rules, has been further developed method for diagnosing computer systems using fuzzy logic, which, unlike the existing uses and description of the classification criteria of computer systems in the form of fuzzy variables that made it possible to significantly reduce the number of production rules and increase the speed of the expert diagnostic system. Scientific and practical results of the thesis used by LLC "Aldek - CCC" (Kharkiv) to diagnose the technical state of a specialized computer cluster, which is evidenced by a certificate from 1.02.2011r. With the introduction of complexity reduction was achieved, reducing the time cost of search failures in the computer cluster, as well as improved procedures for diagnosis, ensuring appropriate depth troubleshooting by using the intellectual tools of diagnosis.