Design of Technology for Prediction and Control System Based on Artificial Immune Systems and the Multi-agent Platform JADE

Abstract

The work is devoted to the creation of a technology and an intelligent system for prediction and control of complex nonlinear dynamic objects in the oil and gas industry based on modified algorithms of artificial immune systems using the multi-agent platform JADE. As an example, there is considered a technological process for gas purification from acidic components at the U300 installation (medium pressure absorber D302) of the Tengiz Chevroil enterprise. There has been developed a knowledge base based on ontological models implemented in the Protege ontology editor. The creation of the knowledge base is carried out according to the concept of MDA (Model-Driven Architecture), when first platform-independent ontological models are created, and then platform-dependent ones. The hierarchical structure of classes and the visualization of a common ontological model for a multi-agent system are presented. Developed a block diagram of a multi-agent intelligent prediction and control system, implemented on the JADE platform. A description of the containers created in JADE and a fragment of the agent specification are given. Simulation results are presented.