Simulation research of automatic railway air brake system based on neural network

The air transmission characteristics of the railway air brake system of 10 000 t heavy haul train were studied by analyzing the railway air brake test results, the characteristics that affect the key components of the railway air brake system(train pipes, auxiliary reservoirs, and brake cylinders) were extracted, and a simulation model of automatic railway air brake system based on neural network was established. A simulation model of an automatic air braking system based on neural networks was established. The model was input into the longitudinal dynamics model as a braking excitation, and the obtained results were compared with the actual operation results of a 10 000 t heavy haul train on the Hao−Ji Railway. The results show that the braking and release signals of a 10 000 t heavy haul train are transmitted from the locomotive to the remote vehicle. With the increase of the transmission distance, the transmission speed is almost constant, but the transmission intensity is attenuated. The brake system simulation model based on the neural network can predict the changes of the air pressure of the train pipes, auxiliary reservoirs, and brake cylinders under the condition of 50 kPa common brake pressure reduction for 10 000 t heavy haul trains with a prediction accuracy of 99.9%. Compared with the traditional fluid dynamics simulation models, the calculation efficiency is improved about 2 938 times and has wide engineering application prospect. © 2024 Central South University of Technology. All rights reserved.