Abstract:With the rapid development of the subway, passenger satisfaction has become an important indicator for measuring the brand value of subway services. Based on the Consumer Brand Equity (CBBE) model framework, the random forest model was used to systematically explore the quantitative influence mechanism of brand-building elements on passenger satisfaction. Data was collected through a questionnaire survey of Chengdu Metro, and empirical analysis was conducted by distinguishing three groups: commuters, tourists, and elderly passengers. The research finds that the brand demands of three types of passengers show significant differentiation characteristics, and there is a significant synergy effect between the punctuality of trains and the promotion of low-carbon and environmental protection for commuters. The identification of tourists" characteristic theme stations has a reinforcing effect on the cultural publicity effect. The interaction between the LOGO recognition of elderly passengers and the service of flight attendants can significantly enhance the perceived experience. It provides empirical support for the application of CBBE theory in transportation scenarios and also offers a decision-making basis for metro operators to formulate differentiated brand-building strategies.

Abstract:To provide a theoretical basis for operational safety risk assessment of heavy-haul trains, this study investigates common power-loss faults in the braking system through multi-condition simulations based on a longitudinally dynamic model validated by experiments. Using nine typical fault modes and seven typical line segments derived from actual operational scenarios, the maximum tensile compressive coupler forces and longitudinal acceleration responses under fault conditions were analyzed. The simulation results from the established model closely align with experimental data, demonstrating the model’s reliability. The findings indicate that the level of longitudinal impulse is collectively determined by the fault type, dynamic state, and line gradient. Specifically, power loss in the non-operating section of the leading locomotive under traction (Fault?1) generated the maximum tensile coupler force of 1,370?kN on a 3‰ upgrade. Power loss in the operating section of the trailing locomotive (Fault?3) produced the highest peak longitudinal acceleration of 7.59?m/s2 on a 4‰ upgrade. Power loss in the non-operating section of the trailing locomotive under combined braking and traction conditions (Fault?7) resulted in the maximum compressive coupler force, exceeding 1,000?kN, on a descending section with a gradient transition from -12‰ to -8‰. Moreover, faults occurring at gradient transition zones further intensify longitudinal impulses and elevate operational risks. This research provides theoretical and data-driven support for the development of fault emergency strategies and operational optimization.

Abstract:The identification of key path chains serves as the foundation for optimizing traffic control strategies. Based on trajectory data from partial vehicles in a mixed traffic flow environment, this paper proposes a key path chain identification method that integrates maximum a posteriori estimation with multi-metric fusion. First, leveraging road network topology and connected vehicle trajectory data, a joint identification model for path set classification is constructed using maximum a posteriori estimation theory, enabling the recognition of key path chain sets between different node pairs in the network. On this basis, considering the effects of traffic congestion hysteresis, path divergence characteristics, and dynamic flow fluctuations, three metrics—path flow aggregation capacity, path discreteness, and operational impedance—are proposed. A nonlinear key path chain set criticality evaluation and ranking model is established by integrating these three metrics. Simulation verification is conducted using road network and traffic survey data from a specific area in Nanchang. The rationality of the identification results is tested using path traffic transfer intensity, and the sensitivity of the model under different penetration rates is analyzed, demonstrating the effectiveness of the proposed model.

Abstract:To accurately evaluate the forces and deformations of adjacent pipelines induced by shield tunnel construction, a Pasternak-TB mechanical model with piecewise uniform parameters is established. First, the pipeline is modeled as a Timoshenko beam on a Pasternak foundation. This model is then analyzed using a two-stage method to determine the forces and deformations induced by shield tunnel excavation. In the first stage, the additional load on the axis of the existing pipeline induced by tunnel excavation is calculated. In the second stage, a Timoshenko beam model on a Pasternak foundation with piecewise uniform parameters is established to simulate the pipeline. By applying the obtained load to this mechanical model, an analytical method for determining the pipeline forces and deformations is derived. Subsequently, the proposed method was validated against field data, existing theoretical methods, and centrifuge tests. Finally, the influence patterns of ground loss ratio and pipeline elastic modulus on pipeline forces and deformations were analyzed. Results indicate that the pipeline"s maximum displacement and bending moment increase linearly with the ground loss ratio. In contrast, an increase in the elastic modulus nonlinearly reduces the maximum displacement but increases the maximum bending moment. The proposed method provides an efficient tool for tunnel-pipeline interaction analysis; the revealed patterns offer a theoretical basis for ground loss ratio control and differential verification of pipeline strength.

Abstract:Ball k-means, as an accurate acceleration algorithm for k-means, improves efficiency while guaranteeing consistent clustering results. However, its acceleration effect diminishes on high?dimensional data, and the neighbor?searching step involves high complexity. To enhance its adaptability to high?dimensional data while preserving its acceleration advantages in large?k clustering, this paper proposes an optimized algorithm named Ball k?means?G*. In the data point assignment step, a primal?geometry structure is introduced by projecting data points and candidate centroids into two?dimensional space, where a lower bound for distances is established to skip redundant high?dimensional computations. In the neighbor?cluster search, a forced?point mechanism is incorporated: centroids are reduced to three dimensions, and combined with the original algorithm’s pruning conditions to eliminate non?neighbor distance calculations early. Experiments on real datasets with different scales and dimensions show that the proposed algorithm significantly outperforms five existing exact accelerated k?means methods. Compared with the original Ball k?means, it achieves an average efficiency improvement of approximately 20.66% across datasets of various dimensions. The results demonstrate that Ball k?means?G* effectively alleviates the performance degradation of Ball k?means in high?dimensional settings and is applicable to clustering tasks across various data dimensions and with large k values.

Abstract:Coatings prepared by laser cladding technology possess excellent properties such as wear resistance, corrosion resistance, and resistance to fatigue wear, along with the advantage of achieving metallurgical bonding between the coating and the substrate. This study aims to realize the cost-effective and efficient repair of localized railway wheel damage using laser additive technology, employing three different types of alloy powders: Fe-based, Co-based, and 316L stainless steel. Localized repairs were carried out on wheel surfaces, and the rolling contact fatigue performance of the three coatings was comparatively analyzed. ER9 wheel steel was machined into small-scale wheels with a diameter of 60 mm, and notches were introduced on the surface to simulate localized damage. Using a TF-YF6000 laser system, Fe-based, Co-based, and 316L stainless steel coatings were deposited on the substrate surface under the following parameters: laser power of 2600 W, spot diameter of 3 mm, scanning speed of 0.6 m/min, and an overlap ratio of 50%. Rolling-sliding friction tests were subsequently conducted on the LGPS-30C wheel–rail contact simulation test bench. The microstructural morphology, phase composition, and nanohardness of the coatings were analyzed using a scanning electron microscope (SEM), X-ray diffractometer (XRD), optical microscope (OM), and nanoindenter, respectively. The results indicated that the coating surfaces were dense and exhibited good metallurgical bonding. The Fe-based, Co-based, and 316L stainless steel coatings primarily consisted of dendritic and eutectic microstructures, and their hardness was significantly enhanced compared to the base wheel material. Macro- and microscopic analysis of the wear morphology revealed critical differences in performance among the three coatings: the Co-based alloy coating exhibited a notably smooth and flat wear surface without any signs of crack initiation, demonstrating the best wear resistance; in contrast, the 316L stainless steel alloy coating showed significant flaking and more severe wear; while the Fe-based alloy coating displayed a relatively flat surface, it developed deeper cracks and extensive crack propagation at the substrate–coating interface, indicating a potential risk of coating fracture and detachment. The surface of the Fe-based coating exhibited fine scratches and ploughing marks, with an abrasive wear mechanism. The Co-based coating surface showed evidence of material accumulation, indicating an adhesive wear mechanism. The 316L stainless steel surface displayed pronounced spalling, with its wear mechanism primarily characterized as fatigue wear.

Abstract:To address the limited imaging resolution and depth-quantification accuracy in the detection of hidden defects within the rail head, an ultrasonic phased-array detection and imaging-optimization procedure for rail-head defects is proposed. A two-dimensional acoustic-field model of the rail head is established in COMSOL to investigate the influence of array parameters on the focused beam distribution and to optimize the probe-parameter set. Inspection data are acquired using Full Matrix Capture (FMC) and reconstructed by the Total Focusing Method (TFM), and an imaging-optimization strategy combining normalization-based denoising, the Hilbert transform, and TF-PCF weighting is introduced. Experiments are conducted on a P60 rail test block with prefabricated flat-bottom-hole defects in the rail head, and the results are validated by comparison with simulations. The results show that improved imaging performance is obtained with 32 elements, an element-pitch-to-wavelength ratio of 0.5, an element-width-to-pitch ratio below 0.5, and a steering angle not exceeding 45°. The optimized imaging suppresses noise and artifacts and enables defect localization and quantitative measurement, with a hole-length error within 10% and a depth error within 5%. This study provides a theoretical basis for high-resolution imaging and quantitative evaluation of rail-head defects in ultrasonic phased-array inspection.

Abstract:In urban rail transit systems, Communication-Based Train Control (CBTC) utilizes bidirectional wireless communication to enable real-time data exchange between trains and trackside equipment, thereby ensuring normal train operations. However, the open operating environment of CBTC exposes it to the threat of cyber attacks. To address this issue, this paper proposes a collaborative intrusion detection method based on reputation weight and trusted distribution. This approach first employs differential privacy and secret sharing techniques to support locally trained detection models, dynamically evaluating node reliability through subjective logic. Subsequently, it combines a reputation-weighted aggregation algorithm to effectively suppress malicious attacks and enhance system stability. Finally, blockchain technology is introduced to establish a trusted distribution mechanism, ensuring secure model updates. Simulation experiments on the CBTCset dataset demonstrate that the proposed method achieves an accuracy rate of up to 99.6%, outperforming traditional average-weighted aggregation and privacy-preserving methods in terms of accuracy, F1, precision, recall and time delay.

Abstract:Single-image super-resolution is a challenging ill-posed problem. Current methods based on convolutional neural networks face performance bottlenecks, while Transformer models, though capable of improving performance through global modeling, struggle to achieve computational efficiency due to their high computational complexity. Therefore, we propose a progressive Cross-Learning Network (CLNet) that integrates ultra-dense dilated residual blocks (UD2B) with enhanced Transformer blocks (ETB) to construct a synergistic progressive architecture. UD2B aggregates high- and low-frequency features through multi-scale dilated convolutions to enhance local representations, while ETB establishes long-range dependencies via cross-channel self-attention to capture global context. We also introduce a Cross-Feature and Cross-Level Attention Fusion Block (C2AFB) that achieves effective fusion of multi-level features through adaptive learning. Experiments on multiple benchmark datasets demonstrate that CLNet outperforms existing methods in both objective metrics and visual perceptual quality, achieving a favorable balance between performance and efficiency.

Abstract:To study the face stability of a shield tunnel drilling underneath an existing tunnel at close distance, a three-dimensional failure mechanism of the soil in front of the excavation face was constructed by using spatial discretization technology. Utilizing this failure mechanism and virtual work principle, the objective function of the chamber pressure for the shield machine under limit states was derived. The upper bound solution of the required chamber pressure for maintaining the stability of soil in front of the tunnel face during was obtained based on optimization calculations. Moreover, the influences of various parameters on both the chamber pressure and the failure modes of the front soil mass were investigated. Parametric analysis demonstrates that the required chamber pressure decreases with the shield machine approaches the existing tunnel and increases with it moves away from the existing tunnel. Applying the proposed theoretical methodology to the Guangzhou Metro Line 12 shield tunneling project, the theoretical solution of the chamber pressure were calculated. By comparing with the field-measured data of the chamber pressure, the validation of the proposed approach was proved.