Research on Detection and Correction of High-Speed Train Idling and Skidding Based on Improved Fuzzy Neural Network

In order to solve the problem of weak idling and skidding detection in high-speed train operation, an improved adaptive fuzzy neural network method for detecting and correcting idling and skidding of a high-speed train is put forward in this paper. This paper simulated the output speed data of 4-speed sensors of the high-speed train, designed an improved adaptive fuzzy neural network, selected some standard operation data and idling and skidding data, and trained and tested the fuzzy neural network, according to the detection results, the data of nonidling and nonskidding were fused by the adaptive weighted average method. The accelerometer sensor was used to compensate for the idling and skidding time. The simulation results show that the improved algorithm has a more extensive detection range and higher accuracy than the fixed threshold method. The correction design can effectively compensate for the train speed. The method is proved effective in detecting weak idling and slip, improving the speed measurement accuracy of high-speed trains, and having a specific application reference value.