Intelligent Information Support for Decision Making in Maintenance and Equipment Repair Management

The concept of organization of intellectual information support for decision-making, including primary processing of data on failures and symptoms, is discussed. Determination of probabilities and verification of the most probable symptom with recalculation of values of all posterior probabilities is considered in this article. The formalization of knowledge and formation of production system rules in the form of recommendations are shown. Also authors discuss organizing decision-making information support using symptom-based failure detection technology. As well as classification of equipment and components based on key factors in order to identify the most "in-demand" components and components taking into account coefficients associated with class at planning purchases in order to ensure timely supply of components are discussed. The solution is made using Bayes method based on statistical information on relationship of signs with states and on frequency of manifestation of these states. In the process of knowledge formalization, authors proposed to use production system rules in procurement planning, classification of components/nodes using neuro-fuzzy systems, forecasting and optimization methods. New knowledge obtained in the process of data analysis allows expanding the information base. Practical realization is performed using analytical MATLAB platform and EXSYS Corvid developed in USATU.