Data Fusion and Industrial Equipment Diagnostics Based on Information Technology

Abstract

Heterogeneous data fusion obtained in real time from several sources is a key objective when diagnosing equipment in all industries and transport. The paper proposes a heterogeneous data fusion model and a model for equipment diagnostics which are based on the levels of the Joint Directors of Laboratories (JDL) model using Data Mining technology; ontology storage containing clear, fuzzy and fuzzy-functional cognitive models; soft computing methods and probabilistic-static methods. The models proposed consider the heterogeneous data fusion in terms of a systemic approach and are not tied to any industry. A detailed description of the forecasting model of electrical equipment (EE) at an oil company is considered. The models developed will improve the accuracy and completeness of EE diagnostics through joint processing by expanding the volume of data received, taking into account the knowledge of staff on duty which will in turn lead to the timely identification of EE faulty states and failures, the equipment repair, reduction of the time spent by a chief engineer for decision-making to repair or replace the EE.