Neural Network Approach to Railway Stand Lateral Skew Control

TL;DR

This paper explores neural network-based modeling and control methods for a railway stand's lateral skew, addressing challenges faced by traditional model-based control through real-data neural network approaches.

Contribution

It investigates neural network architectures and training methods for real-data modeling and control of railway stand lateral skew, highlighting their potential advantages.

Findings

Neural networks effectively model the railway stand's lateral skew.

Real-time training methods are suitable for adaptive control.

Experimental results demonstrate the feasibility of neural network control.

Abstract

The paper presents a study of an adaptive approach to lateral skew control for an experimental railway stand. The preliminary experiments with the real experimental railway stand and simulations with its 3-D mechanical model, indicates difficulties of model-based control of the device. Thus, use of neural networks for identification and control of lateral skew shall be investigated. This paper focuses on real-data based modeling of the railway stand by various neural network models, i.e; linear neural unit and quadratic neural unit architectures. Furthermore, training methods of these neural architectures as such, real-time-recurrent-learning and a variation of back-propagation-through-time are examined, accompanied by a discussion of the produced experimental results.