Formalizing natural parallelism of management and information processes

Abstract

In modern automated control systems for complex objects (ACS CO), the processes of obtaining, storing, processing and transmitting data, information and knowledge are closely intertwined with the processes of monitoring and managing these objects. At the same time, these processes can proceed both sequentially and in parallel, using various resources of the ACS CO. The analysis shows that within the framework of existing approaches, when describing the processes under consideration, the dynamic nature of the partial order that arises between them is not taken into account. Such a dynamically changing partial order in previous publications is proposed to be called the natural parallelism of the processes of functioning of the ACS CO. In these papers, it is shown that with automatic purposeful dynamic parallelization of processes, high efficiency is achieved in the use of the resulting models in the control of complex objects. The implemented parallelism is close to the natural parallelism of the simulated processes. At the same time, the execution of processes using dynamic parallelization is distinguished by the possibility of scaling to a different number of parallel interpreters (processors, parallel threads). In this article, a possible approach to formal description of the natural parallelism of management and information processes is proposed.