Teplynskyi K. Parallel computer simulation environment, oriented to parameters identification of complex dynamic systems

Dissertation is devoted to research and improve the organization of computer resources, focused on the identification and optimization of parameters of models of complex dynamic systems, such as biological and chemical models. It is shown the role of the parallel simulation environment as a form of the organization of computing and information resources for complex dynamic systems. The structure of the parallel simulation environment is designed, that focuses on modeling, optimization and parameter identification of complex dynamic systems. The requirements for parallel simulation environment, which are defined on the basis of its functions and simulation stages. The decomposition of hardware and software is implemented for distributed parallel simulation environment with regard to the identification of parameters for complex dynamic systems. Developed optimization subsystem based on genetic algorithm modifications. The testing mechanism implemented for all parameters. Designed optimization subsystem is integrated into the parallel simulation environment DIANA, which is developed by the Max-Planck Institute (Magdeburg, Germany) together with the Donetsk National Technical University, and has been successfully used to solve problems of global optimization of complex dynamic system.