The dissertation is devoted to the solution of the actual scientific and technical problems, consisting of the substantiation, development, and study of computer models of the system for diagnosing the technical condition of electric motors (mining machines aimed at increasing the objectivity of diagnostics of electromechanical equipment using methods of statistical evaluation, Petri nets and cluster analysis).
The need for scientific work is due to the fact that in modern conditions at mining enterprises there are questions of increasing the reliability and efficiency of electromechanical equipment (EME), first of all, electric motors (EM) of mining machines. A significant part of used equipment needs replacement due to exhaustion of the limited service life or overhaul. As a result, material and time costs increase due to equipment downtime, product damage caused by accidents, reduced electrical and fire safety, abnormal modes of operation, and overload voltages greater than normal, and other undesirable consequences. In view of the complexity of non-destructive diagnostics of the technical conditions for electromechanical equipment, or even the full absence of such
diagnostics, the costs of repairs increase, as well as the amount of repair work that needs to be done. However, an analysis of the experience of using EME shows that a significant part of it still has a sufficiently large margin of reliability and, if there are reasonable
recommendations, the duration of the life cycle of such equipment can be significantly increased. Such recommendations can be obtained by computer modeling of diagnostic systems for technical conditions of electric motors of mining machines. Due to this, the dissertation research is devoted to the development of effective computer models of diagnostic systems of the technical conditions for electric motors, which will allow providing early detection of defective conditions, to increase the reliability of EME.
The presented work substantiates the choice of the set of basic energy-mechanical parameters, which have an impact on the technical condition of mine electric motors, and it is formed a system of classification of modes of technical condition of EM provided that they are divided into four appropriate classes, including normal operation, current, and capital repairs and а complete failure. Computer models are proposed to determine the dependences between leakage current through interphase insulation and phase currents, as well as between current and voltage phase shift and leakage insulation of electric motors, which allow using these circumstances for current control and the prevention of insulation defects and accident connected with its violation.
Mathematical models for diagnosing the technical condition of electric machines of mining machines based on the Bayesian statistical method, Petri nets, and cluster analysis have been developed using Kohonen neural networks, adaptive resonance, and extreme machine learning. The use of such models will increase the objectivity of diagnostics of the technical conditions of electromechanical equipment by conducting uninterrupted monitoring of energy-mechanical parameters in real-time. This will ensure the elimination of the shortcomings of the existing system of planned and preventive repairs in the formation of repair schedules for effective organization of the maintenance under the actual conditions and reliable operation of both, old and new types of equipment.
Keywords: computer modeling, diagnostic system, technical condition, electric
motors of mining machines, statistical methods, Petri nets, cluster analysis
