The excitation and vibration triggered by the long-term operation of railway vehicles inevitably result in defective states of catenary support devices. With the massive construction of high-speed electrified railways, automatic defect detection of diverse and plentiful fasteners on the catenary support device is of great significance for operation safety and cost reduction. Nowadays, the catenary support devices are periodically captured by the cameras mounted on the inspection vehicles during the night, but the inspection still mostly relies on human visual interpretation. To reduce the human involvement, this paper proposes a novel vision-based method that applies the deep convolutional neural networks (DCNNs) in the defect detection of the fasteners. Our system cascades three DCNN-based detection stages in a coarse-to-fine manner, including two detectors to sequentially localize the cantilever joints and their fasteners and a classifier to diagnose the fasteners' defects. Extensive experiments and comparisons of the defect detection of catenary support devices along the Wuhan-Guangzhou high-speed railway line indicate that the system can achieve a high detection rate with good adaptation and robustness in complex environments.

This paper aims to develop a combination method for the classification of power quality complex disturbances based on ensemble empirical mode decomposition (EEMD) and multilabel learning. EEMD is adopted to extract the features of complex disturbances, which is more suitable to the nonstationary signal processing. Rank wavelet support vector machine (rank-WSVM) is proposed to apply in the classification of complex disturbances. First, the characteristic quantities of complex disturbances are obtained with EEMD through defining standard energy differences of each intrinsic mode function. Second, after the optimization of rank-SVM, based on wavelet kernel function, the ranking function, and multilabel function are, respectively, constructed. Lastly, rank-WSVM is applied to classify the complex disturbances. Simulation results and real-time digital simulator tests show that for different signal to noise ratio, the rank-WSVM classification performance of complex disturbances including hamming loss, ranking loss, one-error, coverage, and average precision, is generally better than the other three methods, namely rank-SVM, multilabel naive Bayes, and multilabel learning with backpropagation.

The increased complexity and intelligence of automation systems require the development of intelligent fault diagnosis (IFD) methodologies. By relying on the concept of a suspected space, this study develops explainable data-driven IFD approaches for nonlinear dynamic systems. More specifically, we parameterize nonlinear systems through a generalized kernel representation for system modeling and the associated fault diagnosis. An important result obtained is a unified form of kernel representations, applicable to both unsupervised and supervised learning. More importantly, through a rigorous theoretical analysis, we discover the existence of a bridge (i.e., a bijective mapping) between some supervised and unsupervised learning-based entities. Notably, the designed IFD approaches achieve the same performance with the use of this bridge. In order to have a better understanding of the results obtained, both unsupervised and supervised neural networks are chosen as the learning tools to identify the generalized kernel representations and design the IFD schemes; an invertible neural network is then employed to build the bridge between them. This article is a perspective article, whose contribution lies in proposing and formalizing the fundamental concepts for explainable intelligent learning methods, contributing to system modeling and data-driven IFD designs for nonlinear dynamic systems.

Split pins (SPs) play an important role in fixing joint components on catenary support devices (CSDs) of high-speed railway. The occurrence of loose and missing defects of SPs could make the structure of CSDs unstable. In this paper, we present a three-stage automatic defect inspection system for SPs mainly based on an improved deep convolutional neural network (CNN), which is called PVANET++. First, SPs are localized by PVANET++ and the Hough transform & Chan-Vese model, and then, three proposed criteria are applied to detect defects of SPs. In PVANET++, a new anchor mechanism is applied to produce suitable candidate boxes for objects, and multiple hidden layer features are combined to construct discriminative hyperfeatures. The performance of PVANET++ and several recent state-of-the-art deep CNNs is compared in a data set that is collected from a 60-km rail line. The results show that our model is superior to others in accuracy, and has a considerable speed.

The utilization of a large number of power electronic converters in the high-speed railway system results in the low-frequency instability for the single-phase vehicle-grid system. In order to probe into the problem, the impedance-based approach is adopted. This paper focuses on the dq-frame impedance modeling and verification of the single-phase converters in the electric multiple units. The closed-loop input impedance modeling methods for both single converter and interleaved converters considering the second-order generalized integrator-based phase-locked loop (PLL) are put forward. Then, the dq-frame impedance measurement method for single-phase systems based on the Hilbert transform is proposed, and the modeling approach is verified by the frequency-domain simulations on the switching models. The good agreement of the model and simulation illustrates that the modeling method is suitable and accurate enough until half of the switching frequency. It is elucidated that the single-phase rectifier's impedance has the off-diagonal characteristic and the negative-impedance characteristic on the dd channel at low frequencies. Moreover, both the current and voltage controls have influences on the negative impedance shaping, while PLL does not.

The wind-induced vibration of the high-speed catenary and the dynamic behaviour of the pantograph–catenary under stochastic wind field are firstly analysed. The catenary model is established based on nonlinear cable and truss elements, which can fully describe the nonlinearity of each wire and the initial configuration. The model of the aerodynamic forces acting on the messenger/contact wire is deduced by considering the effect of the vertical and horizontal fluctuating winds. The vertical and horizontal fluctuating winds are simulated by employing the Davenport and Panofsky spectrums, respectively. The aerodynamic coefficients of the contact/messenger wire are calculated through computational fluid dynamics. The wind-induced vibration response of catenary is analysed with different wind speeds and angles. Its frequency-domain characteristics are discussed using Auto Regression model. Finally, a pantograph model is introduced and the contact force of the pantograph–catenary under stochastic wind is studied. The results show that both the wind speed and the attack angle exert a significant effect on the wind-induced vibration. The existence of the groove on the contact wire cross-section leads to a significant change of the aerodynamic coefficient, which affects largely the aerodynamic forces applied on the catenary wires, as well as the vibration response. The vibration frequency with high spectral power mainly concentrates on the predominant frequency of the fluctuating wind and the natural frequency of catenary. The increase in the wind speed results in a significant deterioration of the current collection. The numerical example shows that a relatively stable current collection can be ensured when the wind flows at the relatively horizontal direction.