Algorithms for Intelligent Data Processing in Resource Management Under Uncertainty and Dynamic Environment

Abstract

The article discusses algorithms for intelligent data processing that improve the efficiency of solving resource management problems in conditions of uncertainty and dynamism of the external environment. As a fundamental information technology that provides collection, storage, retrieval, processing and use of data and knowledge for resource management in conditions of uncertainty and semantic constraints of the subject area, data processing technology using a multi-agent approach has been chosen. The solution to the problem in the form of a resource allocation plan is built as a result of agents negotiating with each other. Agents participating in the negotiation process determine the admissibility and acceptability of a solution for them based on an individual knowledge base. The possibility of self-learning of agents of a multi-agent system is laid down at the level of individual knowledge bases belonging to individual agents. New knowledge gained as a result of negotiation in solving the problem of resource management, as well as based on the experience gained, update the knowledge base of agents in the form of appropriate rules and restrictions, which can be used in the future to improve their efficiency. The algorithms were tested on the example of solving the resource management problem in production scheduling.