| Popis: |
Examples of building models and measures in monitoring and diagnostics of electric power objects are given. It is known that accuracy and reliability of results of diagnostics of technical objects depends on many factors. And, not in the last place, from qualitatively formed training sets which on different parameters and characteristics correspond to certain technical conditions of investigated objects. The questions connected with the appearance of some physical processes and their mathematical models accompanying the work of electric power equipment units are considered. The results of mathematical models formation of training sets (measures) which correspond to different technical conditions and modes of robots of the investigated electric power equipment are given. The choice of diagnostic spaces, the coordinates of which are the estimates of parameters or functional characteristics of diagnostic signals, is justified. Known in statistics scattering ellipses are used as learning sets, the boundaries of which with a certain probability cover the data of the results of experiments obtained on real power equipment. A scheme and an algorithm implementing it are proposed, which allow to form learning sets that take into account both possible types of defects in individual nodes of electrical power equipment and their modes of operation (rotor speed of the electric machine, temperature of the diagnosed node, various degrees of electrodynamic and mechanical loading, etc.). This approach allows to use monitoring and diagnostics systems within the Smart Grid technology, which provides the possibility to diagnose power equipment in real time. For building training sets that correspond to both certain types of defects and modes of operation of power equipment units, the results of experimental studies obtained at the laboratory stands of the Institute of Electrodynamics of the National Academy of Sciences of Ukraine were used. As an example of practical application of the proposed models, the problem of constructing solving rules at vibrodiagnostics of rolling bearings of electric machines has been considered. |
