One of the important tasks of maintaining the rolling stock in operating condition is the diagnostics of axle-box units with roller bearings, allowing timely fault detection in all bearing elements. For technical condition assessment of the rolling stock axle-box bearings, recognition algorithm that allows to compare a posteriori information with a priori class description by the diagnostic features language was given in the paper. As a result of diagnostics, it was found that axle-box bearings, when making the diagnosis, may have type I and type II errors. If the feature actually belongs to the serviceable condition class, and the diagnostics system decides on the presence of defects in the rolling bearing, then there is a type I error. If the decision on serviceable condition is accepted when the roller bearing is faulty, then there is a type II error. Obviously, the result of these errors is different, so different costs are appointed to them. Therefore, the cost of risk during diagnostics is appointed by an expert operator. Author Biography Василь Григорович Равлюк, Ukrainian State Academy оf Railway Transport Feuerbach Square 7, Kharkov, Ukraine, 61050 PhD, Senior Lecturer Department of Wagons
Alan : Fen Bilimleri ve Matematik
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|