Currently, the detection equipment for offline arcing on high-speed trains is of great interest because it can evacuate the contact state between the contact wire and the strip. In order to ensure the safety of the train under the effect of the arcing, the self-designed fourth-order Hilbert curve fractal antenna is chosen to detect the arcing, in this paper, based on its electromagnetic radiation signal. Aspects of time-, frequency-, and joint time–frequency-domain analyses of the signals were considered. Meanwhile, the experiments were conducted several times with different circuit currents and different electrode gaps, and the electromagnetic pulses under different moments were studied. Results indicated that the electromagnetic radiation pulses were corresponding to the abrupt variations of voltage during arcing, and all the characteristic frequencies were located at similar values (18 MHz). Mechanisms of using the short-time Fourier transform method to obtain an accurate arcing moment, extinguished time, and arcing duration have been explained in detail. Thus, the characteristics of arcing electromagnetic radiation could become the significant and efficient parameters to detect the pantograph–catenary operation state of high-speed trains.

When a high-speed train approaches an insulated split-phase section embedded between neighbouring power supply arms carrying powers with different phases, the vehicle-mounted vacuum circuit breaker (VCB) needs to be switched off to avoid short-circuit fault during the phase-changing procedure. At the moment of switching the VCB contacts, the arc is triggered easily, accompanying with overvoltage impulse. As the core of the whole system, the traction transformer frequently suffers the invasion of the overvoltage impulse, which possibly causes partial discharge or breakdown accidents. To evaluate the impact of overvoltage on the transformer, it is essential to observe the overvoltage distribution at different positions in the transformer. As an equivalent circuit model of the traction transformer involving the spatially distributed capacitance is built based on the improved capacitor network methodology (ICNM), the partial discharge risk in the transformer has been assessed. Ultimately, the multi-physics coupling models based on the finite element method have been launched for observe the specific distribution of the multiphysics field for verifying the validity of equivalent circuit model based on ICNM. These works lay the theoretical foundation for guiding the design of insulation and the thermal dissipation settings of the traction transformer, considering the threat brought from impulse.

The performance of the pantograph–catenary system plays a significant role in determining the reliability and safety of a high-speed train. With an increase in train speed, repeated separations between the pantograph and catenary result in an increased number of arcing events, which exacerbate the levels of damage on the pantograph–catenary system, leading to frequent train accidents. The arcing event is a complex phenomenon that is influenced by various factors, and involves interactions between the electromagnetic, thermal and airflow fields. The high temperatures generated by an arc result in ablation of material in the pantograph–catenary system, thus reducing its life expectancy. In order to investigate the mechanism of this damage to the pantograph–catenary system, a model of the arcing phenomenon is established and analyzed. In the proposed model, a set of arc plasma conservation equations are solved to determine the temperature distribution in the arc. The influences of different experimental conditions on the characteristics of the arc and the temperature distributions in the catenary wire and pantograph strip are calculated and discussed. Finally, the results obtained in the simulation studies are compared with realistic arc images recorded using a high-speed camera, and a good agreement is found.

The sliding electric contact of pantograph/contact wire systems plays a significant part in the current collection stability and operation life of pantograph/contact wire systems. This paper addresses the evolutionary process of sliding electric contact of pantograph/contact wire systems by analyzing three key characteristics including contact resistance, temperature distribution and microstructure. The influence of electric current on contact resistance was interpreted. Furthermore, the evolution of the spatial temperature distribution of the carbon strip was obtained in the zigzag movement, while the dominant role of electric effect in temperature rise was demonstrated. In the end, the wear morphology differences under pure friction and current-carrying conditions were analyzed. The formation of radial cracks was illustrated and its influenced on the wear process was discussed.

Composite insulators are prone to accelerated aging in coastal and industrially polluted environments, leading to flashovers, power grid outages, and economic losses. Traditional detection methods either require power grid shutdown or lack adequate evaluation capabilities. This research introduces a pixel-level assessment of the aging status of composite insulators using hyperspectral imaging (HSI) technology. And proposed a least squares support vector machine (LSSVM) based on Improved Beluga Whale Optimization (IBWO) algorithm to evaluate the aged levels of composite insulators. First, artificial accelerated aged samples are categorized into aging levels I-VI using the static contact angle. Hyperspectral data in the 400 nm–1040 nm wavelength range are acquired using a HSI device, followed by preprocessing steps involving denoising and dimensionality reduction. Subsequently, the IBWO algorithm is employed to identify the optimal parameters ( C , σ ) for LSSVM. The performance of IBWO was compared with other optimization algorithms on test functions, demonstrating that IBWO exhibits the optimal convergence speed and optimization accuracy, effectively enhancing the classification and generalization ability of LSSVM. In this study, the proposed method was compared with other algorithms using k-fold cross validation, resulting in overall accuracy of 96.83% and demonstrating its superior classification capability. The presented method enables pixel-level assessment of the aging degree and visualizes the distribution of aging states in composite insulators. It provides valuable guidance for studying the aging characteristics and structural design in diverse complex environments, holding significant potential for noncontact evaluation of other electrical equipment.

The cable is an essential pivot of electric energy transmission, and its operational condition influences the safety and stability of both the power system (PS) and traction power supply system (TPSS). The prompt detection of cable insulation condition by partial discharge monitoring is crucial for mitigating potential damage to the cables. To begin with, this article provides a comprehensive overview of partial discharge detection approaches for cables by introducing the mechanism of partial discharge in cables. The detection approaches are categorized into electrical detection techniques and non-electrical detection techniques. Afterwards, to accurately assess the insulation condition of cables using the partial discharge detection method, this article summarizes the insulation defect diagnosis approaches for cables. In particular, the insulation defect diagnosis approaches primarily comprise extraction and optimization for features, traditional machine-learning-based fault diagnosis approaches, and deep-learning-based fault diagnosis approaches. To conclude, this article summarizes the challenges and future research directions of cable fault diagnosis. Accurate cable fault diagnosis is fundamental to maintaining the reliable operation of cables, ensuring an uninterrupted power supply to both PS and TPSS, and enhancing the responsiveness to equipment failures.