Scenario analysis of unconventional emergency disaster accidents based on the knowledge elements and dynamic Bayesian networks
LI Jian-han · Applied Mechanics and Materials · 2014
This article intends to provide a scenario analysis of unconventional emergency disaster accidents based on knowledge elements and dynamic Bayesian networks. The so-called knowledge elements we mean to say may include the Case,the Disaster,the Object,the Activitywhen they are used to represent scenario of some unconventional emergency disaster accidents. It is on basis of introducing related theories of Dynamic Bayesian Networks that we resort to all network node variables and set thresholds. On condition that node variables are greater than threshold value,they should be considered to be key knowledge elements that tend to affect catastrophe,and vice versa. However,if key knowledge element that influences said disaster,it is likely to be applied to build up evolution network of scenarios of unconventional emergency disasters,it would also be possible to be used to work out in-situ probability of each node variable by utilizing joint probability formula so as to achieve a scenario deduction. The present paper also wants to quote July-16 Dalian oil depot fire as a case study sample to analyze deduction in hoping to demonstrate specific process of scenario evolution analysis model. The result of our analysis shows that: probability of pipeline bursting and blowout is about 72% at 18: 12,whereas possibility of T103 tank bursting and blowout symbolized a96% at 18: 19,and,in same way,the pump bursting and blowout is 85% at 21: 30. The result is just in correspondence with actual scenario in-situ,which proves effectiveness and feasibility of this method. It is just because of complex,diverse and uncertain characteristic features of influencing factors of unconventional emergency disaster accident,the key information elements may symbolize higher demands on sample information. Therefore,it is only when sample information and data is enough representative that sufficiently experienced experts in field can predict probability of accidents more accurately through logic deduction.