Adaptive multi-agent planning of industrial resources based on ontology
Abstract
Introduction: Currently, new solutions are required in managing industrial resources, in order to maintain a high level of adaptability and efficiency. Classical combinatorial or heuristic methods and tools cannot provide adequate solutions for real-time resource management. Purpose: Development of a method for planning industrial resources based on multi-agent technologies and ontology, in order to adapt the system to unforeseen events, such as new orders, unavailable resources, etc. Results: An adaptive planning method has been developed, in which agents continuously improve the system performance in real time by identifying and resolving conflict situations caused by unforeseen events. To adjust multi-agent planning to specific features of the production process, semantic networks (ontologies) are used, which are the basis of ontological knowledge bases for storing information about the peculiarities of a particular enterprise. In this regard, the following elements have been developed: the basic planning ontology, an ontology editor for creating a specialized enterprise ontology, a knowledge base in the form of a semantic Wikipedia for the enterprise, and a multiagent scheduler which can be customized using the basic and specialized ontologies in accordance with specific production features and requirements for the technological operations. Practical relevance: Application of the system developed with the method of planning industrial resources is not limited to machine-building enterprises, but can be recommended for managing projects, supply chains, etc.Published
2018-12-20
How to Cite
Skobelev, P., Lakhin, O., Mayorov, I., & Simonova, E. (2018). Adaptive multi-agent planning of industrial resources based on ontology. Information and Control Systems, (6), 105-117. https://doi.org/10.31799/1684-8853-2018-6-105-117
Issue
Section
Control in social and economic systems