Fuzzy Inference Model for Railway Track Buckling Prediction

The application of rail buckling models is often limited by uncertain information with respect to track properties, and many conventional are poorly suited network-wide or even regional application. Here, a methodology using fuzzy sets presented that, when trained data can use inputs properties predict the minimum temperature increase for particular track. An investigation impact size training influence key parameters on was conducted, it found that high level stems from sleeper spacing fastener torsional resistance parameters. model shown give low prediction error small dataset sizes data. results this work show efficacy based applied data, giving both rapid calculation times. approach has potential wider array variables, such as geometry vehicle dynamics, not study owing flexibility inference methodology.