Lepikhin A.M.  

Modeling and computing technologies in problems of risk analysis of complex engineering systems

Complex high power systems and engine systems are the basis of modern industrial infrastructure. Most of them are critical systems, which crash leads to hard consequences for the economy of regions and countries. Study of the critical systems safety is one of the major scientific challenges of our time. Safety concept is based on the risk analysis methodology.
The risk analysis methodology has three key tasks. The first task is to analyze the limit states of complex systems, which achievement leads to structure disaster. The second task is to forecast the probability of limit states achievements and disasters. The third task is to identify the critical elements of the structure, which damage or destruction may inevitably lead to disaster.
The first task requires the development of limit states models for system structure. This is not a trivial challenge because it needs to consider a number of damage mechanisms and limit states criteria, and also to take into account the possibility of joint-damage process. The difficulty in solving the second problem is connected with necessity of probabilistic descriptions of limit states parameters. You need to consider the properties of random processes and fields, and also the theory of random emissions. The third task requires constructing models of complex hierarchical structures survivability, as only survival analysis can identify the critical elements. Multi-level synergistic models of complex structures damage can be the solution of this problem.
Creating of informative and compact risk analysis models with a large number of significant parameters is one of the most difficult tasks that requires the use of modern computing technologies such as parallel computing, GRID-technology and Cloud-technology. Parallel computing is effective for solving multiparameter problems of limit state analysis of complex systems. GRID-technologies enable to describe and analyze multiparameter random processes and fields. Cloud-technologies allow to analyze the distribution of information databases on the technical system state, external influences, characteristics of engineering materials, safety requirements, and to create information databases for decision-makers.