ANALYSIS OF METHODS AND TOOLS OF ARTIFICIAL INTELLIGENCE SYSTEM DEVELOPMENT

Abstract

PURPOSE. The paper provides the analysis of tools and methods of new expert system development using expert system shells. METHODS. The operation process of automated control systems of process plants requires constant estimation of the values of various parameters. It is often when these parameter values can not be obtained by traditional methods involving direct measurement or calculation by known analytical dependencies. This situation is especially typical for the systems with a high level of uncertainty which do not have a sufficient past history of operation. Heuristic methods based on the assessments of experts in the relevant field are appropriate in this case. RESULTS AND THEIR DISCUSSION. The obtained results of the study have allowed to systematize the knowledge on the application of the development methods of new expert systems using expert system (ES) shells which take into account the features of different programming tools and provide higher efficiency through the exhaustive search for more alternatives when choosing a solution. CONCLUSIONS. Based on the high-quality experience of a group of specialists the development methods using ES shells analyze the effect of a larger volume of new factors through their evaluation under the construction of strategies and addition of prediction capabilities.