Implementation of Wavelet Algorithm and Maximum Change-Point Method for the Detection of Ballast Substructure Using GPR

Abstract

A geophysical method named ground-penetrating radar (GPR) can be applied to geotechnical investigation to determine the underground conditions. In this study, a GPR is used to detect the ballast condition of the railway track. The ballast bed plays an important role in rail track keeping the gauge between sleepers and thereby position of the rails. As the train moves over the track, ballast helps to hold the track in place. Ballast fouling is formed by coal dust and the breakdown of ballast or from soil interference under the rail track. The rail track can be damaged substantially due to the fouling material such as coal, metal piece and large stone. This paper presents the non-destructive testing method, namely ground-penetrating radar (GPR) for identification of different types of materials. To detect the underground substructures, 800 MHz ground-coupled antenna was used. Constructed a model rail track with different fouling material are buried under the subsurface structures. This work is carried out in Civil Engineering laboratory, IISc, Bangalore. This research work presents the detection of underground structures like different metals, big-sized ballast and clean ballast. Find maximum change-point algorithm is proposed to detect the abrupt changes in the signal; by this, it will also detect the size of a target by calculating the start and end trace of the buried object. It is important for the diagnosis of ground substructures. This paper also describes the application of wavelet decomposition to detect the condition of the ballast viz clean or fouled. The wavelet decomposition finds the shape of the object by calculating the higher-order statistical parameters like skewness test and kurtosis test.