Analytical Risks Prediction. Rationale of System Preventive Measures for Solving Quality and Safety Problems

Abstract

The approach for building new probabilistic models which is intended to predict risks for complex intellectual structures under different threats and to solve quality and safety problems in system life cycle, is proposed. The approach includes: selection of models for every system element allowing to estimate probability to lose integrity during given prognostic period; the approach to build probability distribution functions (PDF) by the models selected to predict probability and risks to lose integrity for different scenarios of threats; the approach to integrate PDF for structure from two elements united by serial and/or parallel connection and generalization for complex structure; the approach to form input for modeling and to solve quality and safety problems in application to complex intellectual structures; applications, which can prove efficiency in different areas. The approach is applicable for the analysis of the reliability of complex systems built from unreliable components, estimations of increasing expected reliability and safety for intelligent manufacturing, modeling of robotic and automated systems operating in cosmic space, optimization of a centralized heat supply system, estimating the mean residual time before the next parameters abnormalities for monitored critical systems, control of information quality, estimation of human factor, analysis of vulnerability of sea oil and gas systems etc. Effects are demonstrated by practical examples.