Gennady A. Oparin, Vera G. Bogdanova, Anton A. Pashinin

Matrosov Institute for System Dynamics and Control Theory SB RAS

The proposed technology is based on methods, models, software tools, and automation tools for solving problems of qualitative research of binary dynamic systems based on the Boolean constraint method in microservice computing infrastructure. The essence of the method is to reduce the problems under consideration to solving SAT or 2QBF problems by constructing a Boolean model of a dynamic property. The software includes tools for building Boolean models and checking their feasibility. We implemented these software tools as a package of applied microservices as part of our technology. The launch of the microservice is controlled by software agents installed in the nodes of the computing infrastructure, created based on the previously developed HPCSOMAS-MSC instrumental platform. This study aims to extend HPCSOMAS-MSC with tools for automation of computing nodes preparation in a distributed environment for deploying a package of applied microservices and specialized agents focused on the structure of a Boolean model to provide two-level control of parallel computing when checking its feasibility. At the upper level, the subtasks obtained by splitting the Boolean model are distributed over the nodes of the lower level, subjected to deep splitting, and solved in parallel as independent ones. Computational experiments for a qualitative study of binary dynamic systems with a large state vector dimension have shown the effectiveness of the proposed technology. The wide use of such systems as object models in conducting research in various subject areas determines the practical significance of the proposed technology.

binary dynamic systems, qualitative analysis, Boolean constraint method, microservice