Development of Smart-Technology for Forecasting Technical State of Equipment Based on Modified Particle Swarm Algorithms and Immune-Network Modeling

Abstract

The article is devoted to the development of Smart-technology for forecasting technical state of industrial equipment based on artificial intelligence methods. One of the most important tasks in forecasting is the creation an optimal set of descriptors that most fully characterize industrial equipment’s work. Preliminary data processing and the selection of informative descriptors based on modified particle swarm algorithms have been performed. The application of modified particle swarm algorithms allows to investigate a search space in more detail and to avoid premature convergence. The forecasting technical state of equipment and image recognition have been carried out based on immune-network modeling. The developed Smart-technology is used to forecast the technical state of equipment based on real-life production data of TengizShevroil oil and gas company. The modeling results have been obtained on the basis of daily measurements from industrial Installation 300.