2010Unpublished venueRequires access

A Novel Algorithm for Predicting Track Irregularities of Unit Track Sections

Renwei Liu, Peng Xu, Qiang Sun

Open publisher page 0 citations

Abstract

During train runs, the interaction between train wheels and the rail track underneath makes track geometry change, which in turn results in all kinds of track irregularities. After the 6th train speed raise of China in 2007, railway transportation has shown three main new features: speed-raised, heavy-loading and high-density. Under these features, changes in railway track irregularities of China have also presented some new characteristics: higher deterioration rates of track irregularities and more frequent occurrences of track exceptions. To ensure the train operational safety and increase the transportation service quality, the preventive inspection and maintenance of railway track facilities have been put forward once again by railway maintenance departments of China. A precondition for the preventive inspection and maintenance is about how to accurately evaluate and predict the future track condition according to the historical track inspection data. In this paper, based on the characteristics of track irregularity changes and in accordance with the calculus thinking, we have developed a short-range prediction model called SRPM. The model uses track waveform data generated by the track geometry car (TGC) to predict track irregularities of a unit track section with the length of 100m for each day in a future short period of time. An algorithm for using SRPM to predict track irregularities has also been designed. According to the designed algorithm, using ORACLE database and computer program languages, we have programmed a computer software named P-SRPM. We then used P-SRPM to deal with 25 sets of TGC-generated track waveform data from the up going track of the Beijing-Shanghai railway (Jing-Hu railway) administrated by Jinan Railway Bureau (JRB) and predicted track irregularities of unit sections in the railway track segment. Finally, errors in these predictions were analyzed in both temporal and spatial dimensions. From the error analysis results, we come to the conclusion that SRPM can fairly accurately make short-range predictions for track irregularities of each unit section in the JRB-administrated Jing-Hu railway track (up going).

About this research paper

What this paper is about

During train runs, the interaction between train wheels and the rail track underneath makes track geometry change, which in turn results in all kinds of track irregularities. After the 6th train speed raise of China in 2007, railway transportation has shown three main new features: speed-raised, heavy-loading and high-density. Under these features, changes in railway track irregularities of China have also presented some new characteristics: higher deterioration rates of track irregularities and more frequent occurrences of track exceptions. To ensure the train operational safety and increase the transportation service quality, the preventive inspection and maintenance of railway track facilities have been put forward once again by railway maintenance departments of China. A precondition for the preventive inspection and maintenance is about how to accurately evaluate and predict the future track condition according to the historical track inspection data. In this paper, based on the characteristics of track irregularity changes and in accordance with the calculus thinking, we have developed a short-range prediction model called SRPM. The model uses track waveform data generated by the track geometry car (TGC) to predict track irregularities of a unit track section with the length of 100m for each day in a future short period of time. An algorithm for using SRPM to predict track irregularities has also been designed. According to the designed algorithm, using ORACLE database and computer program languages, we have programmed a computer software named P-SRPM. We then used P-SRPM to deal with 25 sets of TGC-generated track waveform data from the up going track of the Beijing-Shanghai railway (Jing-Hu railway) administrated by Jinan Railway Bureau (JRB) and predicted track irregularities of unit sections in the railway track segment. Finally, errors in these predictions were analyzed in both temporal and spatial dimensions. From the error analysis results, we come to the conclusion that SRPM can fairly accurately make short-range predictions for track irregularities of each unit section in the JRB-administrated Jing-Hu railway track (up going).

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

During train runs, the interaction between train wheels and the rail track underneath makes track geometry change, which in turn results in all kinds of track irregularities. After the 6th train speed raise of China in 2007, railway transportation has shown three main new features: speed-raised, heavy-loading and high-density. Under these features, changes in railway track irregularities of China have also presented some new characteristics: higher deterioration rates of track irregularities and more frequent occurrences of track exceptions. To ensure the train operational safety and increase the transportation service quality, the preventive inspection and maintenance of railway track facilities have been put forward once again by railway maintenance departments of China. A precondition for the preventive inspection and maintenance is about how to accurately evaluate and predict the future track condition according to the historical track inspection data. In this paper, based on the characteristics of track irregularity changes and in accordance with the calculus thinking, we have developed a short-range prediction model called SRPM. The model uses track waveform data generated by the track geometry car (TGC) to predict track irregularities of a unit track section with the length of 100m for each day in a future short period of time. An algorithm for using SRPM to predict track irregularities has also been designed. According to the designed algorithm, using ORACLE database and computer program languages, we have programmed a computer software named P-SRPM. We then used P-SRPM to deal with 25 sets of TGC-generated track waveform data from the up going track of the Beijing-Shanghai railway (Jing-Hu railway) administrated by Jinan Railway Bureau (JRB) and predicted track irregularities of unit sections in the railway track segment. Finally, errors in these predictions were analyzed in both temporal and spatial dimensions. From the error analysis results, we come to the conclusion that SRPM can fairly accurately make short-range predictions for track irregularities of each unit section in the JRB-administrated Jing-Hu railway track (up going).

Key concepts: Track (disk drive), Track geometry, Computer science, Algorithm, Waveform, Telecommunications, Operating system, Radar

Related papers

Back to paper searchBrowse research topicsOriginal source
A Novel Algorithm for Predicting Track Irregularities of Unit Track Sections — Research Paper | ScholarLens