Study on the Vibration Behaviour of Locomotives Under Wheel Polygon Excitation and Its Quantitative Detection Method

Wheel polygons can significantly impair the operational quality, service life and ride comfort of railway systems and even threaten the safety of train operations. Timely detection of wheel polygons is of great benefit in formulating a reasonable wheel maintenance strategy. First, to investigate the vibration response of the locomotive under wheel polygon excitation, a locomotive–tack coupled dynamic model is established, taking into account the flexibility of the bogie frame, wheelset, sleeper and track structure. The effects of typical wheel polygons on the vibration behavior of the locomotive under different conditions were studied. Then, the simulation results were used to quantitatively analyze the correlation between the axle box acceleration (ABA) of the locomotive and the wheel-rail system P2 mode, the inherent modal of the wheelset, the vehicle speed, the order and the amplitude of the wheel polygons. Besides, a dataset with multiple working conditions was created from the simulation data and a quantitative wheel polygon detection model was constructed based on the deep learning algorithm. Finally, the effectiveness of the proposed detection method was verified using the real-world data of ABA and wheel polygons. The results show that the proposed method can quickly and accurately detect the dominant features of the wheel polygon, i.e. the wavelength and the corresponding roughness information.