USE OF NEURAL-NEUTRAL LOGICAL SYSTEMS FOR EVALUATING THE TECHNICAL CONDITION OF THE MAINLINE LOCOMOTIVE DIESELS ON THE BASIS OF THE PARAMETERS OF THE ENGINE OIL

The article focuses on the use of Sugeno-type adaptive neuro-fuzzy inference systems (ANFIS) for the assessment of the technical condition of mainline diesel locomotive engines, based on the motor oil tribodiagnostic parameters, without engine disassembly. The main engine units, which include the cylinder-piston group, valve-injector group, plain bearings and fuel dilution of oil, are equipped with subsystems designed to be independent of one another. In addition, a comparative study of trapezoidal and Gaussian membership functions for the input variables is given, and the results obtained are demonstrated to be reliable.

Hali tarjima qilinmagan