Seeking Rare Events in a Simulated System using Risk Distances

Gollücke, Volker and Hahn, Axel
SIMUL 2017, The Ninth International Conference on Advances in System Simulation
The identification and avoidance of potential risks is an important task in any system design process. The probability of underestimating a risk depends strongly on the complexity of the system under consideration. It is helpful to know the ways and means, which help to review the suspected risks. Current safety regulations, for example in the domain of offshore operations, require a description of all the risks involved. In order to create this description, a system including a behavioral specification modeled by a system expert is analyzed. Within this system, risks have to be searched and reasons for their occurrence have to be analyzed. The approach presented in this work describes a method to calculate the distance to a risk using a simulative analysis. The approach uses concepts from the field of Rare Event Simulations, which are used to accelerate the simulative reach of a risky situation. During the simulation runs data is collected about the way, which was taken to reach the avoidable situation. These data can be further used in a manual risk analysis and for matching rules and processes in terms of safety. The presented work describes how distance functions are created and used to assess system states in terms of their proximity to risky situations. In addition, the question of how the results of the distance functions can be used to guide co-simulations to the examined risky situations is answered.
Oktober / 2017
978-1-61208-594-4 / 2308-4537
Think Mind
Interdisciplinary Research Center on Critical Systems Engineering for Socio-Technical Systems