This research presents an approach based on artificial intelligence techniques for wheel polygonization detection. The proposed methodology is tested with dynamic responses induced on the track by passing a Laagrss-type rail vehicle. The dynamic response is attained considering the application of a train-track interaction model that simulates the passage of the train over a set of accelerometers installed on the rail and sleepers. This study, which considers an unsupervised methodology, aims to compare the performance of two feature extraction techniques, namely the Autoregressive Exogenous (ARX) model and Continuous Wavelets Transform (CWT). The extracted features are then submitted to data normalization considering the Principal Component Analysis (PCA) applied to suppress environmental and operational effects. Next to data normalization, data fusion using Mahalanobis distance is performed to enhance the sensitivity to the recognition of defective wheels. Finally, an outlier analysis is employed to distinguish a healthy wheel from a defective one. Moreover, sensitivity analysis is performed to analyze the influence of the number of sensors and their location on the accuracy of the wheel defect detection system.
Polygonal wheels are one of the most common defects in train wheels, causing a reduction in comfort levels for passengers and a higher degradation of vehicle and track components. With the aim of contributing to the safety and reliability of railway transport, this paper presents the development of an innovative methodology for classifying polygonal wheels based on a wayside system. To achieve that, a numerical train-track interaction model was adopted to simulate the passage of a freight train over a virtual wayside monitoring system composed of a set of accelerometers installed on the rails. Then, the acquired acceleration time series was transformed to a frequency domain using a Fast Fourier transform (FFT), and on this data, damage-sensitive features were extracted. The features based on Principal Component Analysis (PCA) showed great sensitivity to the harmonic order, while the ones based on Continuous Wavelet Transform (CWT) model showed great sensitivity to the defect amplitude. One step further, all features are merged using the Mahalanobis distance in order to obtain a damage index strongly correlated with the polygonal defect. Finally, a cluster analysis allowed the automatic classification of polygonal wheels, according to the harmonic order (harmonic-based) and defect amplitude (amplitude-based). The proposed methodology demonstrated high efficiency in identifying different types of polygonal wheels using a minimum layout of two sensors.
This study presents a comparative analysis of three AutoEncoder (AE) models—Variational AutoEncoder (VAE), Sparse AutoEncoder (SAE), and Convolutional AutoEncoder (CAE)—to detect and quantify structural anomalies in railway vehicle wheels, such as polygonization. Vertical acceleration data from a virtual wayside monitoring system serve as input for training the AE models, which are coupled with Hotelling’s T2 Control Charts to differentiate normal and abnormal railway component behaviors. The results indicate that the SAE-T2 model outperforms its counterparts, achieving 16.67% higher accuracy than the CAE-T2 model in identifying distinct structural conditions, although with a 35.78% higher computational cost. Conversely, the VAE-T2 model is outperformed in 100% of the analyzed scenarios when compared to SAE-T2 in identifying distinct structural conditions while also exhibiting a 21.97% higher average computational cost. Across all scenarios, the SAE-T2 methodology consistently provided better classifications of wheel damage, showing its capability to extract relevant features from dynamic signals for Structural Health Monitoring (SHM) applications. These findings highlight SAE’s potential as an interesting tool for predictive maintenance, offering improved efficiency and safety in railway operations.
Artificial-intelligence-driven wayside monitoring has become a promising solution for early identification of railway wheel flats, enabling safer operations and more efficient maintenance planning. This study introduces a comparative investigation of supervised and unsupervised machine learning strategies for wheel flat identification, with particular emphasis on real-time applicability and sensor cost reduction. Support Vector Machines (SVMs) and k-means clustering are evaluated as representative supervised and unsupervised approaches using vibration data obtained from numerically simulated train–track interactions under realistic operating conditions, including train speeds of 120 km/h and 200 km/h and multiple wheel flat severities. A key contribution of this work is the proposal of a simplified supervised classification framework that directly exploits Auto-Regressive features extracted from rail-mounted accelerometers, eliminating the need for feature normalization and multi-sensor data fusion. This simplification significantly reduces computational effort, making the approach suitable for real-time deployment in operational railway environments. In parallel, a systematic sensitivity analysis is conducted to assess the influence of sensor placement and to identify the minimum sensor configuration required to achieve reliable damage classification. The outputs from the current study show that an SVM emerges with more accurate defect classification than the k-means clustering, allowing a wayside system with fewer sensors.