Situational Modelling of Reliability and Safety for Industrial-Natural Systems
Keywords:
Situational Modelling, Logic Model of Reliability, Logic Model of Safety, Triggering Event, Logic Probabilistic MethodAbstract
Introduction: The situational approach to modelling the state of industrial-natural systems (INS) and to comparing alternative
structures of their realization proposed earlier by one of the authors is generalized for the problems of developing logical models of
structural reliability and safety of such systems. The research is aimed at developing the technique to build logical models of structural
reliability and safety of INS functioning on basis of their normal functioning model. Results: Dangerous and critical situations are
modeled as an "extension" of normal performance models, using the same software environment. The modelling process is detailed
down to a single material or informational signal transferred between the system elements. This is the main difference of the proposed
technique from the conventional ways of modeling safety/reliability of complex systems on basis of their structure charts. It provides
the following advantages: the opportunity to specify and formalize descriptions of functionally, spatially and temporally generated
failures typical for INS as well as triggering events and conditions whose appearance and development depends on spatial-temporal
attributes of INS elements; the principle possibility to estimate partial operability of an INS at failures of certain elements taking into
account the current quantity of the resources and rates of their spending in order to calculate acceptable restoration times for failed
elements and analyze the consequences of their failures; the support of building and studying alternative logical models during INS
construction, as well as the systems to control its safety and reliability. Practical relevance: This technique can raise the validity of
logical models of safety and reliability by more complete consideration of interconnections between INS elements, thus revealing the
most dangerous situations and scenarios caused by combined and multiple failures within the system. A common modeling environment for both normal and emergency modes of INS functioning provides accumulation of diverse kinds of knowledge about the investigated object for their integration.