An Intelligent Wheelset Spinning Detection System in a Direct Current Traction Drive

Abstract

A spinning detection method of a locomotive wheelset in a dc traction drive and its implementation on the basis of intelligent processing of measuring data of the drive control system is considered. An increase in the amplitudes of the low-frequency harmonics of the voltage sensor output signal on the locomotive power circuit elements is used as a sign of loss of motion stability of the wheelset due to reaching the threshold wheel–rail adhesion. For automatic multilevel identification of spinning modes, a random forest algorithm-based classifier is used. The results of the analysis of the control power circuit parameters preindicated by experts show the high sensitivity and accuracy of the proposed intelligent spinning detection system.