A neural network approach for railway safety prediction

Abstract

Edgbaston, B15 2TT, UK Salford, A45 4wT, UK S.Nefti-Meziani&alford. ac. uk M.Oussalah@bham.ac.uk Abstract -Artificial Neural Networks (AMVs) are becoming increasingly popular for solving complex problems, as they can behave quite well at solving proLlems that don't have an algorithmic solution or for which the algorithmic solution is too complex to be found. In railway systems, the problem of predicting the sJ1steni malfinctions. or equivalently, railway safety is ofparamount interest for most of railway companies. Traditional ways of predicting railway safe@ are very expensive in terms of time consuming, which make them inefficient under certain circumstances. This paper advocates the use of AMVs architecture to handle the safety problem. By taking irregularities in the positioning of the rails as input to the ANN the ANNcanpredict the safety ratio of the rails.In order to reduce the dimensionality of inputs data a wavelet transformation technique has been employed. Drfferent neural network structures are created and lheir performances both in terms of mean squared error and correlation coeficient have been evaluated to$nd om the best structure for predicting railway safety. The experiments showed that when the model is trained cw a dataset subset and then tested on different subser. it perjormed satisfactorily and con predict the desired output with a very low error factor.

Wireless Sensor Networks: Toward Smarter Railway Stations

“…The regression algorithms using machine-learning techniques for model prediction has been conducted to investigate the vertical-acceleration of railway wagons [75]. Moreover, artificial neural networks (ANNs) have been utilized to predict the derailment of a train on the track [76], and more techniques have been applied in the field such as: support vector machines (SVMs) Chong et al (2005), kernel density estimation (KDE) by Anderson (2009), and clustering by Lee et al (2004) [77][78][79].…”

A neural network approach for determination of Preisach model parameters under a sinusoidal induction at various frequencies

“…NN gains these weights by training, which means that an ANN must be trained before it can contain useful knowledge. One of the main algorithms used in training is the back propagation algorithm originally developed by Werbos [4,5] and [6]. The basic idea behind back propagation algorithm is that the output of neurons in the lower layer is sent to the upper layer until they reach the output layer.…”

Reduction of power consumption in sensor network applications using machine learning techniques

“…Wireless sensor networks are widely used to monitor railway tracks and irregularities, detect abandoned objects in railway stations and develop intrusion detection systems, secure railway operations, monitor tunnels [13][14][15]. Machine learning techniques have been introduced in different research projects to predict the typical dynamic behavior of railway wagons running on the track [16][17][18][19][20]. Raw data collection, data pre-processing, and formatting are essential parts of developing any monitoring systems including the above mentioned research works.…”