Abstract Railway track defects must be detected and repaired to ensure the safe operation of the railway. Signals generated by wheelset contact with rail discontinuities contain information about railway track structure and health. Each signal has a characteristic pattern, which is modified in the presence of defects. Signal decomposition is commonly used to extract characteristics from impact response signals. It is necessary to select an appropriate decomposition method for the wheel/rail impact signal. This paper presents a study of the effectiveness of decomposition methods to identify the most appropriate technique for classification of rail conditions. Vibration signals have been generated in experiments and decomposed. The signal characteristics are presented and energy‐based methods are used in an evaluation of the different approaches. Standard deviation is used to assess the similarity of parameter value from different intrinsic mode functions (IMFs); maximum and minimum differences from the mean value are used to consider the effectiveness of each technique in identifying specific damage types. The evaluation shows that variational mode decomposition (VMD) is suitable for the rail application. Classifiers are applied to the outputs of the selected decomposition method. The results demonstrate the suitability of the approach for processing wheel/rail impact response signals to detect faults and improve safety.

Railway track geometry varies along routes depending on topographical, operational and safety constraints. Tracks are prone to degrade over time due to various factors, with deviations from the original geometry design having potential implications for comfort and safety. Regular inspections are carried out to evaluate track condition and determine whether maintenance interventions should be undertaken to correct track geometry. The dynamic measurement of track geometry parameters generates large volumes of data that must be analysed to evaluate track degradation. This work comprehensively explains how track quality is evaluated, introducing four main categories of factors affecting it. These are track design, loading, environment and maintenance. The most common techniques applied to evaluate track condition and predict degradation and faults, categorised into statistical, Machine Learning, Big Data and other, are also introduced. Specifically, the influence of each factor on track geometry is stated and the common techniques applied to each factor determined from this review. The utility of loading and maintenance data for fault prediction depend on the availability of records, whilst the impact of environmental conditions is expected to become increasingly important due to climate change. Artificial Neural Networks, Bayesian models and regression are the most applied techniques for determining track degradation behaviour and fault prediction, considering several different factors in their models. Increasingly sophisticated algorithms can consider multiple factors in tandem to predict faults based on the unique conditions of specified tracks.

Inspection of physical surface defects is a significant concern in many industrial areas. In railway systems, this process mainly includes the detection and classification of defects in rails and wheels, for which laser‐based optical inspection technologies have gradually been applied in the form of 2D profile measurement, benefiting from its high precision and robustness to surface conditions. However, defect classification and evaluation after the initial detection works still rely heavily on human inspectors to make maintenance suggestions. The linear nature of rails makes it possible to increase the dimension of rail measurement data from 2D to 3D by aligning 2D profiles along the rail, from which more comprehensive diagnosis information becomes available. In combination with appropriate artificial intelligence algorithms, this approach can potentially replace human‐dominated defect classification and evaluation work. This study presents a 3D model‐based railway track surface defect classification and evaluation method. A set of geometrical features are extracted from the 3D model of track surface defects to describe a distinguishable pattern for each category of defect. Multi‐class classifiers are then tested and have shown promising results on a group of artificial track surface defects, giving a systemic solution for 3D model‐based automatic track surface defect inspection.

There usually exists a high density of traffic through bottleneck sections of mainline railways, where a perturbation of one single train could result in long consequential delays across a number of trains. In the event of disturbances, rescheduling trains approaching the bottleneck will be necessary to increase the throughput of the section. To model the real-time train rescheduling problems around bottleneck sections, a mixed-integer programming model is presented in this paper. An innovative improved algorithm (DE_JRM) is developed to solve the problem. The model and the algorithms are validated with a case study using Monte Carlo methodology, which demonstrates that the proposed algorithm can reduce the weighted average delay and satisfy the requirements of real-time traffic control applications.

The IMSL Fortran library has long supported parallel computing with MPI. As IMSL has progressed to more platforms like C, Java and .NET, our support for MPI has not, but there are tools available to parallelize IMSL solutions on these platforms. These options include ASPEED's ACCELLERANT, Interactive Supercomputing's Star-P, Brain Murmur's JIVA and the Digipede Network. This presentation will present these various options for parallelization of your IMSL-based solutions that aren't written in Fortran.