Simulation-based forecasting effects of an accidental explosion on the road. Part II: Case study

• Main Author: E. R. Vaidogas
• Format: Article
• Language: English
• Published: Vilnius Gediminas Technical University 2006-12-01
• Series: Transport
• Subjects: road accident; collision; explosion; aleatory uncertainty; epistemic uncertainty; simulation; nested loop
• Online Access: http://www.transport.vgtu.lt/upload/tif_zur/2006-4-vaidogas.pdf

The paper contains a case study covering forecasting mechanical effects of an explosion which can be generated during a road accident. It illustrates a practical application of the simulation-based procedure developed for such forecasting in the first part of the paper. The case study reveals the amount and character of the knowledge necessary to carry out this forecasting. Its final result is a probabilistic model describing likelihood of occurrence of accidental explosion as well as characteristics of the incident blast wave generated by this explosion. The accident simulation is based on the classical Bayesian approach to risk assessment. The case study described in the paper shows how to formulate the initial knowledge in line with this approach. Particular attention has been given to handling subjective information (expert opinions) within the problem under analysis. It is shown that this information is indispensable for dealing with the sparseness of hard experience data on most of the phenomena leading to an accidental explosion. The stochastic simulation demonstrated in the paper serves the purpose of propagating uncertainties related to these phenomena. The probabilistic action model describing the potential explosion takes account of these uncertainties.