Analysis and Prediction of Time Series Geometric Parameters of Groove Rail Based on Big Data Technology

In recent years, modern trams have developed rapidly and groove rails are mainly used for tramway tracks. Whether their geometric parameters exceed the limit is related to the safety of train operation. In addition, the health of the grooved rail also affects the service life of the rail equipment. Therefore, track maintenance activities play an important role in maintaining the health of track equipment, extending its service life, and operating safety of railways. With the previous research, we have developed a rail inspection trolley based on the inertial reference method, and it can achieve high-precision detection of geometric parameters of groove rails. On this basis, this paper proposes a set of prediction algorithms based on RBF radial basis neural network. It can accurately predict the change trend of geometric parameters such as track gauge, abrasion, track longitudinal level irregularity and track direction, so as to further analyze the detection data and explore its inherent value, providing a more scientific basis for the preventive track maintenance.