Some Problems of the Simulation Model Efficiency and Flexibility

P. 532–538.

Mikov A., Kozlov A., Zamyatina E., Ermakov S.

Simulation remains one of the most popular and sometimes the only method of investigation of the complex dynamical systems nowadays. The problems become more and more complicate and so simulation models become complicate and large scale too. So the modelers have to reduce the time of simulation run and overall time of the simulation model design, it’s verification, validation and simulation experiment. The authors believe that the problem of efficiency and flexibility of simulation systems (program tools and linguistic construction) can be solved through it’s automation and intellectualization. This paper discusses the problem of design and implementa-tion of the distributed and parallel simulation systems using the power of several compute nodes of computer network (or cluster or multiprocessor computer). Authors suggest original approach to the synchronization and load balancing of the objects of distributed simulation model. Furthermore authors consider the problem of reduction of time needed to design and to debug a simulation model, the problems of the simulation tools adaptability and offer using models and methods of artificial intelligence in these cases

Keywords: ontology, multi-agent systems, parallel and distributed simulation, optimistic and conservative algorithms, automatic simulation model completeness, load balancing

In book

Proceedings of 8th EUROSIM Congress on Modelling and Simulation "EUROSIM 2013», Сarrdiff, Wales, United Kingdom, 10-13 of September, ISBN 978-0-7695-5073-2

Cardiff: IEEE Computer Society, 2013.