United States Department of Transportation
i
Computer Model Developed to Predict Rail Passenger Car Response to Track Geometry
-
2000-10-01
[PDF-109.91 KB]
Details:
-
Alternative Title:Computer model developed to predict rail passenger car response to track geometry : research results
-
Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Corporate Publisher:
-
Abstract:The Federal Railroad Administration sponsored research to develop a computer model to predict the interaction between vehicle and track as a railroad passenger car travels over track with known geometry. This computer model is capable of identifying potentially hazardous sections of track when given the track geometry and the vehicle speed. These predictions will improve safety of railroad operations by helping to determine the maintenance needs for tracks. This computer model, known as a neural network system, estimates the vertical and lateral forces on the wheel/rail interface as a function of the geometry of the track and the operating characteristics of the vehicle. Unlike conventional computer models, a neural network simulates the analytical workings of the human brain. A series of computer models of a railroad passenger car were developed to evaluate the effectiveness of the neural network. This series of computer models accurately represents the dynamic response of an actual railroad passenger car. In Figure 1, the neural network output closely predicts the series of computer models. In the future, the fully developed system will be used to identify track locations where the estimated lateral and vertical forces exceed the limits recommended for safe operations. Only the track geometry and train speed, which are routinely and easily measured parameters, need to be known in order to identify the potentially hazardous locations.
-
Format:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
