CAUSAL NETWORK Vzorová ustanovení

CAUSAL NETWORK. Causal network, which was used to evaluate risks, is illustrated by Fig. 3 and shows the aforementioned subsystems with impact on tunnel safety. The network has seven subsystems (decision-making nodes with numbers 0 through 6), seven random nodes (with numbers 7 through 13) and seven utility nodes (with numbers 15 through 21). It is a Bayesian causal network, see lit. [5] and [6], designed for risk assessment in road tunnels. Network nodes are linked by arrows, which denote their causal relations. Causal relations between individual nodes are defined by the input data (condi- tional probabilities and risk indicators). For every relation, there are causal (parental) nodes and consequent nodes (descendant). Network in Fig.3 includes one original node (0-traffic) without a causal node and seven terminal nodes (all utility nodes 15 through 21) without any consequent nodes. The elaborated analysis con- siders only one risk of injury to health, illustrated by the utility node 21. Relations to other utility nodes (15 to 20) and relations to correspondent nodes are illustrated with regards to the additional expenses due to potential adjustments of existing tunnels or change in existing design documentation. Decision-making nodes represent the level of subsystems and use the value estimated by an expert, based on real condition and the proposed one according to design documentation. Final values, which are illustrated by , are expressed by numbers on scale 1 through 5 :