Search results

Straightness measurement of rail weld joint is of essential importance to railway maintenance. Due to the lack of efficient measurement equipment, there has been limited in-depth research on rail weld joint with a 5-m wavelength range, leaving a significant knowledge gap in this field.

In this study, the authors used the well-established inertial reference method (IR-method), and the state-of-the-art multi-point chord reference method (MCR-method). Two methods have been applied in different types of rail straightness measurement trollies, respectively. These instruments were tested in a high-speed rail section within a certain region of China. The test results were ultimately validated through using traditional straightedge and feeler gauge methods as reference data to evaluate the rail weld joint straightness within the 5-m wavelength range.

The research reveals that IR-method and MCR-method produce reasonably similar measurement results for wavelengths below 1 m. However, MCR-method outperforms IR-method in terms of accuracy for wavelengths exceeding 3 m. Furthermore, it was observed that IR-method, while operating at a slower speed, carries the risk of derailing and is incapable of detecting rail weld joints and low joints within the track.

The research compare two methods’ measurement effects in a longer wavelength range and demonstrate the superiority of MCR-method.

This study focuses on the classification of targets with varying shapes using radar cross section (RCS), which is influenced by the target’s shape. This study aims to develop a robust classification method by considering an incident angle with minor random fluctuations and using a physical optics simulation to generate data sets.

The approach involves several supervised machine learning and classification methods, including traditional algorithms and a deep neural network classifier. It uses histogram-based definitions of the RCS for feature extraction, with an emphasis on resilience against noise in the RCS data. Data enrichment techniques are incorporated, including the use of noise-impacted histogram data sets.

The classification algorithms are extensively evaluated, highlighting their efficacy in feature extraction from RCS histograms. Among the studied algorithms, the K-nearest neighbour is found to be the most accurate of the traditional methods, but it is surpassed in accuracy by a deep learning network classifier. The results demonstrate the robustness of the feature extraction from the RCS histograms, motivated by mm-wave radar applications.

This study presents a novel approach to target classification that extends beyond traditional methods by integrating deep neural networks and focusing on histogram-based methodologies. It also incorporates data enrichment techniques to enhance the analysis, providing a comprehensive perspective for target detection using RCS.