Research on Track Irregularity Full Speed Segment Detection Based on Adaptive Fuzzy Neural Network

In recent years, the construction of China’s rail transportation has grown rapidly. Whether it is an I-shaped track or a grooved-shaped track, the detection of track irregularities is mostly based on the inertial reference method or the string measurement method, but both methods are limited by the speed factor to varying degrees. This paper studies a comprehensive detection method based on fuzzy neural network that combines the inertial reference method with the four-point chord method. It can realize high-precision detection of speed fuzzy speed section (transition speed section), so as to achieve detection of full speed section, and overcome system errors caused by algorithm factors.