Stochastic Hybrid Automaton

• Published in: Systems Dependability Assessment pp. 105 - 120
• Main Authors: Brinzei, Nicolae, Aubry, Jean-François
• Format: Book Chapter
• Language: English
• Published: United States Wiley 2015; John Wiley & Sons, Inc
• Subjects: Monte Carlo simulations; nuclear power plant; steam generator (SG); stochastic hybrid automaton (SHA)
• ISBN: 9781848217652; 184821765X
• DOI: 10.1002/9781119053996.ch7

The stochastic hybrid automaton (SHA) has been implemented in the Scilab/Scicos simulation environment. Scilab/Scicos is an open‐source alternative to commercial packages for dynamical systems modeling and simulation packages such as Matlab/Simulink and MATRIXx/SystemBuild. Scilab includes hundreds of general‐purpose and specialized functions for numerical computation, organized in libraries called toolboxes that cover such areas as linear algebra, numerical integration, simulation, optimization, systems and control, and signal processing. In the context of French research project APPRODYN, the SHA was used to model and probabilistically assess a controlled steam generator (SG) of a nuclear power plant. The behavioral model allows qualitative analysis of system trajectories and identification of dangerous scenarios, even if these are rare and difficult to capture by Monte Carlo simulations. In cases of exponential probability distributions used for times of events, analytical evaluation of a specific scenario probability is also possible.