George Nistratov

Probabilistic modelling processes of mutual monitoring operators actions for transport systems

The research quantitatively proved, that the popular today the mode for estimating risks in operators actions based on uses of exponential probability distribution function (PDF) of time between mistakes of operators is a rough and unpromising engineering way. Its application deforms very essentially probabilistic estimations of risks. Operator may be a man or special device or their combination. For systems, composed from mutual monitoring operators, deviations of risks may be about hundreds and thousands of percent. The more adequate approach for probabilistic modelling processes for is proposed, including intelligent transportation systems, oil&gas transportation systems, infrastructure etc. The logic structure of operators (as a system) and details of threats, system control, possibilities of mutual monitoring and recovery time for every operator are considered. The results are useful for understanding mutual monitoring effects and rationale of counteraction measures against different “human factor” affects. Effects are demonstrated by examples.

Knowledge Mining Based on Applications of The Methods and Technologies of Risks Prediction

This paper is concerned with the knowledge mining approach in application to complex transportation systems. The approach is based on implementing the author’s original mathematical models and supporting them with risk prediction software technologies. Compared to existing methods, the analytically calculated risks depend on the time of periodic control, monitoring, and recovery. Rational use allows the problem to go from data mining according to events to knowledge mining from transportation safety monitoring data and statistics. On the basis of the mined knowledge, proactive decisions may be found before the occurrence of critical events.