Multi-agent Monitoring Information Systems

Abstract

The paper proposes to implement the methodology for the creation of multi-level intelligent monitoring information systems by employing an agent approach with the use of cloud technologies. Information technology of the multilevel intelligent monitoring is implemented in the form of a multi-agent monitoring information system, which is built for the mobile provision of knowledge for each individual decision-making process. Specially created intellectual agents carry out observation and formalization of the emergencies in heterogeneous objects at appropriate levels of monitoring. Functional monitoring aggregation processes carried out in accordance with the global debt arranged by the customer, the genetics of the intelligence agencies of the required classes are studied multiple times for modeling agents and coordination of their interaction, implemented at the upper, server-level management of the multi-agent monitoring information system. The client level contains a multitude of agents and communication tools with them. Using intelligent agents for solving local information retrieval tasks can increase the diversity of the multi-level intelligent monitoring process. It proposed to apply the process of the aggregation of the multi-agent monitoring information system functions at each level of the monitoring and to implement typical units in the form of intellectual agents guided by multi-layered models. Interaction of the intellectual agents has been formed situationally as typical processes of horizontal formation and vertical connections of the method of the ascending synthesis elements of the multilevel structure of the multi-agent monitoring information system. Agents-classifiers, agent-identifiers, predictor agents are identified and described among the agents for data conversion. It is proposed to adopt the method of the synthesis multilayered models to the peculiarities of their use in the structure of the intelligence agents. It was given an example of using one of the methods of the multilayered modeling in the structure of the agent-identifier of the functional dependencies.