Entropia Shannona jako parametr charakteryzujący stan bezpieczeństwa

Niniejszy artykuł jest podsumowaniem wyników badań mających na celu sprawdzenie, czy entropia statystyczna (zwana również entropią Shannona) może być parametrem analitycznym opisującym stan bezpieczeństwa. W pracy dokonano analizy entropii na podstawie danych zebranych przy wykorzystaniu programu EWID dla aglomeracji warszawskiej z lat 1995–2006.

In this article results of analysis which examine the statistical entropy (also called Shannon’s entropy) as an analytical parameter characterizing the state of safety were shown. These results are a summary of studies of entropy made on BW/E – 422/5/2009 project in the Main School of Fire Service. Entropy was calculated from data from the EWID99 program collected since 1995 to 2006 for Warsaw agglomeration. Results of analysis shown that entropy calculated for probability function of casualties (dead or injured people) of fire events is higher for areas characterized by higher fire risk (calculated as a probability and loss ratio). Calculations of Pearson's correlation coefficients for risk and entropy variables shown large positive correlation. This result is an evidence that entropy of probability function of variable which measures the loss, possibly may be used as a parameter which characterizes the state of safety.