High-speed passenger transport and heavy-haul freight transport are the main features of modern railway development. At present, China has built globally leading high-speed and heavy-haul railway networks in terms of scale, and a large number of lines have entered a critical stage of long-term high-load operation and maintenance. How to further enhance the durability of in-service track structures in complex and harsh operating environments and ensure the long-term highly safe, highly stable, and highly reliable service of high-speed and heavy-haul railway track structures has become a major strategic challenge facing the rail transit engineering field in China in the new era. In view of the structural characteristics of high-speed and heavy-haul railway track structures in China and the features of their complex operating environments, to achieve the durability enhancement of track structures, two key scientific issues were condensed (① evolution mechanism of service performance and life assessment and prediction of high-speed and heavy-haul railway track structures and ② durability enhancement mechanism and multi-dimensional regulation mechanism of track structures under complex service conditions), as well as three key technical issues (① high-frequency fatigue characteristics of high-speed railway track components and collaborative regulation technology of wheel-rail system vibration, ② key technologies for component strengthening and performance enhancement of heavy-haul railway track structures under high-traffic-volume conditions, and ③ efficient reinforcement, repair, and rapid retrofitting technologies for track structures). A systematic review of the state-of-the-art progress and cutting-edge advances in the research field was provided. On this basis, the engineering applications and full-scale field validation outcomes of a complete set of durability enhancement technologies, independently developed and deployed by the authors’ team for high-speed and heavy-haul railway lines across typical service conditions were demonstrated. Finally, the future development prospects and critical research priorities in this field were clarified.

With the development of autonomous driving systems, the increasingly prominent bottleneck of “long-tail scenarios” demonstrates that the technology-driven “automation substitution” approach lacks a complete definition of human roles, human-machine relationships, and system ethics. First, the evolutionary pathways of autonomous driving systems from “tool-based assistance” to “automation-led” and ultimately to “collaborative control” were traced, and the fundamental transformation of the human-machine relationship from one-way control to two-way collaboration was elucidated. Secondly, to address key challenges that restrict the improvement of collaborative efficiency, such as the establishment and maintenance of dynamic trust mechanisms, the real-time optimal allocation of control permissions, and the construction of two-way collaborative interaction paradigms, their inherent logic, influencing factors, and existing bottlenecks at the levels of cognitive mechanisms, control theory, and interaction design were expounded on. Finally, developing human state perception and modeling technologies based on multi-source fusion, exploring adaptive collaborative control architectures based on real-time evaluation, and realizing the interaction paradigm shift from “one-way notification” to “two-way collaboration” were systematically proposed, which refined the future development pathways of autonomous driving systems. Through a review of existing theories, technical routes, and application cases, systematic theoretical support and practical references are provided for promoting the development of understandable, trustworthy, and widely acceptable “human-centered collaborative” autonomous driving systems.

4D printing is an innovative technology integrating additive manufacturing with smart materials. Under specific external stimuli, the printed materials can achieve autonomous deformation, which has great application potential in fields such as biomedicine and soft robotics. Its core goal is to realize the dynamic adaptation of structure and function. However, the technology currently faces problems such as poor response synergy of printed materials, difficulty in predicting deformation behavior, high computational costs of structural design, and non-uniqueness of inverse design results. Artificial intelligence provides key support for solving these interdisciplinary complex problems and is the core driving force to promote the intelligent development of 4D printing. The current application status of artificial intelligence in 4D printing was systematically reviewed. The applications of machine learning methods in 4D printing materials, processes, and the forward and inverse designs of structures were emphatically elaborated. The advantages and application potentials of neuro-symbolic artificial intelligence in 4D printing were analyzed, and the practical engineering applications of artificial intelligence-driven 4D printing were summarized. Finally, the remaining challenges in applying artificial intelligence to 4D printing, such as insufficient interpretability, weak generalization ability, and inadequate research on structural fatigue and functional fatigue, were summarized, and future research directions were prospected.

Large tunnel deformation is the most typical engineering geological disaster under complex geological conditions, especially in deep-buried soft rock tunnel engineering. Its occurrence and evolution are controlled by the coupling of multiple factors such as in-situ stress field, structure and mechanical properties of surrounding rock, groundwater activity, and support system response, and it often exhibits progressively accumulative deformation characteristics. Based on Chinese and international studies and typical engineering cases, the definition and classification standards of large tunnel deformation were systematically reviewed. According to dominant hazard-causing mechanisms, large tunnel deformation was classified into four categories: loosening-fracturing type, stress-controlled type, swelling-controlled type, and structure-controlled type, and dominant factors and typical engineering manifestations of each category were clarified. Furthermore, starting from multi-factor coupling mechanisms, key technical paths of construction control were summarized, covering aspects of in-situ stress regulation, surrounding rock improvement, groundwater control, and support system optimization. Two types of support measures, namely resistance-dominated and yielding-dominated ones, were specially summarized. A theoretical basis and engineering references can be provided for mechanism identification, classification recognition, and construction control of large tunnel deformation in complex geological environments.

As tunnel and underground engineering continue to advance toward greater burial depths and increasingly complex environmental conditions, monitoring of the surrounding rock–support system faces growing challenges, including high geostress, multi-physical field coupling, and the strong concealment of structural defects. Conventional monitoring methods have become increasingly inadequate for refined perception, particularly in the identification of localized concealed damage, the capture of dynamic events, and long-term environmental adaptability. Owing to its high-frequency response, active and passive integrated monitoring capability, and good embeddability, piezoelectric sensing technology has gradually expanded from bridges and geotechnical engineering into the field of tunnels and underground engineering. The fundamental mechanisms and monitoring modes of piezoelectric sensing technology were systematically reviewed, as well as its research progress in force and deformation monitoring of surrounding rock–support systems, lining defect identification, dynamic disturbance and disaster-response perception, performance degradation characterization under complex environments, and array-based, distributed, and wireless monitoring. The huge application potential of piezoelectric sensing technology was exhibited in local damage identification and dynamic response monitoring in underground engineering. Future research should focus on piezoelectric response mechanisms and parameter inversion, long-term service reliability under complex environments, array-based and systematic engineering deployment, and intelligent identification methods integrating mechanism constraints with data-driven approaches. Piezoelectric sensing technology is expected to be a key development direction for full-life cycle safety monitoring in intelligent tunnels and underground engineering.

With the frequent occurrence of extreme rainstorms and flood events, water damage of bridges induced by flood impact and scour has become one of the primary factors threatening safe operation. The entire disaster-causing process of bridges induced by flood impact and scour was studied, and the research progress in four aspects, namely scour development mechanisms, structural dynamic responses, intelligent monitoring and early warning, and comprehensive risk assessment, was emphatically discussed and systematically reviewed. From three dimensions of hydrological conditions, structural parameters, and sediment characteristics, the research on the physical mechanisms and evolution laws of local scour around bridge foundations was analyzed. Under the coupling action of hydrodynamic loads and foundation scour, the dynamic response characteristics and typical failure modes of the bridge superstructure, piers and abutments, and foundation system were elucidated. In terms of monitoring and early warning, the monitoring methods based on acoustic, optical, electrical, and mechanical principles were reviewed, and the application potential and current limitations of data-driven and artificial intelligence models in scour depth prediction were emphatically analyzed. At the level of risk assessment, the paradigm evolution from traditional deterministic analysis to probabilistic vulnerability and systemic resilience evaluation was systematically summarized. Based on the shortcomings of existing research, key future research directions were identified, including the scour mechanisms under complex unsteady hydrology and wave-current coupling conditions, the performance evolution of structural systems under the action of multiple disaster chains, the intelligent perception and dynamic early warning of bridges under flood impact and scour based on multi-source information fusion, and the construction of risk and resilience assessment frameworks toward the whole life cycle. Reference can be provided for theoretical research and engineering practice aimed at enhancing the flood impact and scour resilience of bridges.

With the extension of high-speed railways into cold and high-altitude regions, ballastless track concrete structures face severe durability challenges under the long-term coupling action of freeze-thaw cycles and high-frequency train loads. The damage evolution mechanism and research progress of ballastless track concrete under freeze-thaw cycle, fatigue load, and their coupling action were systematically reviewed. Firstly, the cyclic action characteristic of the freeze-thaw cycle was elaborated, and the cross-scale experimental research results of ballastless track concrete were summarized to reveal the damage development law of ballastless track concrete under the freeze-thaw cycle. Furthermore, the statistical characteristics and transfer law of train load were explored. The mechanical properties of the ballastless track concrete structure and the interlayer interface under train load were discussed from the levels of material specimen and full-scale structure tests, and the theoretical analysis framework of ballastless track introducing concrete damage mechanics was generalized. Finally, the synergistic damage effect of freeze-thaw and fatigue coupling action was generalized, and it was pointed out that the coupling action significantly exacerbated the micro-pore development and macro-mechanical property degradation of concrete. The research status and latest progress in this field were reviewed and evaluated, and the existing technical problems in current research and the development trend of future research were clarified to provide theoretical support for the safe operation and maintenance and long-term design of ballastless tracks in cold regions.

Global climate change leads to increasingly frequent extreme low-temperature and severe temperature difference events, intensifying the disaster risk from the instability of natural and engineering slopes in extreme freeze-thaw regions and seriously threatening engineering construction, operation, and maintenance. Under extreme freeze-thaw environments, the slope instability evolution presents complex characteristics of multi-field coupling and multi-factor superposition, which is manifested as a progressive relationship of “freeze-thaw action, hydro-thermal evolution of slope mass, geomaterial degradation, and instability triggering”. The degradation laws of mechanical properties of geomaterials, the hydro-thermal evolution laws of slopes, and their coupling mechanisms with instability modes under freeze-thaw environments were systematically reviewed. The slope stability assessment methods considering freeze-thaw effects were elaborated, including simplified calculation methods and numerical simulation calculation methods. Advances in slope protection and reinforcement technologies suitable for freeze-thaw environments were summarized, with particular attention paid to the mechanical characteristics and design key points of protective structures under frost heave loads. New development ideas in fields such as the calculation of multi-hazard coupling effects and risk assessment of cascading disasters of alpine frozen soil slope instability were discussed, providing engineering references and scientific bases for the construction, operation, and maintenance, as well as risk prevention and control of major projects in severe cold and high-altitude regions.

With the continuous expansion of China’s high-speed railway network and its advancement to the central and western regions with complex geological environments, geological hazards such as land subsidence, landslides, and permafrost pose severe threats to the structural and operational safety of railways. The key technologies and research progress of interferometric synthetic aperture radar (InSAR) in the fine-scale monitoring of geological hazards along high-speed railways were systematically reviewed. Firstly, the mainstream Chinese and international “space-air-ground” SAR systems were sorted out, and the complementary advantages of spaceborne, airborne, and ground-based SAR in the multi-level geological hazard monitoring system of “wide-area general survey, detailed survey of key areas, and local real-time monitoring” were pointed out. Secondly, the key technologies for fine-scale InSAR deformation monitoring to ensure high-speed railway safety were emphatically analyzed, including: deploying artificial corner reflectors (CR) to solve the monitoring difficulties in low-coherence areas; constructing a multi-level coherent target network of permanent scatterer (PS)-distributed scatterer (DS)-CR to achieve high-precision extraction of surface deformation along long-distance railways; utilizing multi-source data fusion to recover multi-dimensional deformation fields, and conducting high spatial-temporal resolution deformation observation through “space-air-ground” collaboration to compensate for the limitations of single SAR orbit observation. Finally, the current challenges faced in aspects such as atmospheric correction under complex environments and disaster early warning were discussed, and the development trend of artificial intelligence empowerment and multi-source integrated perception was prospected, so as to provide key theoretical and technical references for the active safety monitoring and risk prevention and control of high-speed railways.

Driven by the rapid advancement of broadband communications, artificial intelligence, and satellite internet, global data traffic is growing exponentially, while conventional electronic integration is approaching physical limits in bandwidth, power consumption, and latency. Photonics-electronics multiplex integration, leveraging multi-dimensional synergy across the full hierarchy from materials and devices to chips and systems, transcends the bandwidth and power constraints of all-electronic architectures and is emerging as a key enabling technology for next-generation information systems. A systematic review of recent advances in photonics-electronics multiplex integration was presented. Its conceptual framework and developmental trends were first delineated, followed by a survey of high-fidelity simulation technologies that bridge link-level systems and chip-scale implementations, with particular emphasis on the state of the art in heterogeneous integration processes, core devices, and system-on-chip architectures. Representative applications in high-speed communications and intelligent sensing were subsequently examined, followed by a forward-looking discussion of future trajectories toward space, air, and ground integrated networks and AI computing clusters. Three grand challenges that must be addressed were further identified, namely the limits of multi-physics coupling, cross-scale manufacturing consistency, and intelligent adaptive control, and a phased breakthrough roadmap was proposed. These offer strategic insights for ubiquitous 6G coverage and the sustainable evolution of computing networks.

The continuous evolution of intelligent and adaptive threats poses severe challenges to the physical layer security of communication systems. Constructing a secure signal processing mechanism with robustness, predictability, and proactivity is an important foundation for achieving trustworthy communications. To promote the development of generative artificial intelligence (GAI)-empowered secure communication technologies, based on a systematic review of existing studies, the theoretical foundations, key mechanisms, and application progress of GAI in secure communications were analyzed and summarized. First, GAI was formalized as a learning-based signal prior for solving inverse problems at the physical layer. Based on this generative prior, endogenous vulnerabilities could be mitigated; physical layer stability was improved, and a reliable security baseline was established through channel prediction, channel state information completion, and hardware distortion correction. Furthermore, a three-layer proactive defense framework composed of intelligent threat perception, dynamic adversarial game, and covert waveform generation based on noise alignment was constructed. Finally, the key challenges and development directions of GAI-empowered secure communications were prospected, including real-time inference latency, the simulation-to-reality performance gap, the introduction of physics-informed constraints, security digital twins, and the construction of autonomous security agents, and the significance of establishing a unified theoretical perspective and an actionable design paradigm for future research was emphasized.

Bridge structural health monitoring (BSHM) plays a critical role in ensuring the safe operation and extending the service life of bridges. However, the monitoring accuracy and engineering applicability of conventional physics-driven and data-driven BSHM methods are often constrained under the influence of complex operational environments, noise interference, data incompleteness, and model uncertainties. In recent years, physics-informed neural networks (PINNs) and generalized physics-informed machine learning (PIML) methods have developed rapidly, providing new ideas and technical means to overcome the limitations of traditional BSHM methods. The core idea of PINNs is to explicitly or implicitly embed physical prior knowledge, such as physical governing equations and boundary conditions, into deep neural networks, thereby guiding the model to satisfy physical consistency while learning data and improving generalization performance. The theoretical foundations of PINNs/PIML were systematically reviewed, and the advantages and disadvantages of typical physics-embedding strategies, including physical enhancement of feature spaces, physical model data augmentation, physics-informed network regularization, and physics-guided network architecture design, were comparatively analyzed. Focusing on typical tasks in BSHM, such as structural behavior modeling, parameter identification, signal decomposition and reconstruction, as well as damage detection and identification, the latest research advances of PINNs in the field of BSHM were systematically summarized. The main challenges and potential development directions faced by PINNs-based BSHM in practical engineering applications were discussed. With the continuous integration of deep learning methods and physical modeling strategies, PINNs are expected to become an important technical means in intelligent bridge operation and maintenance, providing support for improving bridge condition assessment capabilities and operation and maintenance decision-making levels.

Pantograph slide plates are one of the core components for high-speed trains to obtain electrical energy, and their performance directly affects the current collection quality of the pantograph-catenary system and the operational safety of trains. With the continuous improvement of the operational speed of high-speed railways and the increasing complexity of operational conditions, pantograph slide plates are required to operate stably for a long period under more severe working environments, which imposes higher requirements on their comprehensive performance. The development history of pantograph slide plates was systematically reviewed; the performance advantages and disadvantages of pantograph slide plates at various stages were introduced, and the composition, preparation processes, and research progress of carbon/graphite composite slide plates were summarized. Due to the advantages of light weight, good lubrication performance, and excellent electrical contact stability, carbon/graphite materials have become an important research direction for pantograph slide plates of high-speed trains. However, the further improvement of the performance of current carbon/graphite slide plates still faces many bottlenecks: inherent defects such as pores and microcracks introduced by preparation processes, insufficient interfacial bonding strength between the reinforcement phase and the matrix, and the agglomeration and disordered distribution of the reinforcement phase itself. There are key factors restricting the improvement of their performance. Optimization and surface modification of the reinforcement phase are effective methods to improve the comprehensive performance of carbon/graphite materials. Future research should focus on the design of multi-hybrid reinforcement phases, multi-scale interfacial regulation, and the construction of highly efficient conductive/thermally conductive networks under low filler content, so as to promote the research and development of high-performance carbon/graphite slide plates and their applications in high-speed railways.

With the rapid development of airborne and spaceborne power electronic equipment, the application of pulsed power loads is becoming increasingly widespread in the aerospace field. These loads have strong pulsed characteristics such as high peak-to-average power ratio, wideband variation, and regenerative electric energy, which cause severe impacts on aerospace power supply systems with limited capacity, easily triggering voltage and current fluctuations and even system failures. A systematic review of power conversion technologies for aerospace pulsed power loads was presented. First, the typical power supply architectures and their main features of aeronautical AC/DC power systems and spacecraft electromechanical servo systems were summarized. Subsequently, the structural characteristics and applicable scenarios of pulsed power load suppression topologies such as single-stage, parallel, and cascaded types were classified and summarized. Then, the high-dynamic fast control technologies for different topologies were analyzed, and the core principles and control objectives of various control strategies were elaborated. Finally, the modeling and evaluation methods for stability verification of power supply systems were summarized, and the advantages, limitations, and suitable scenarios of each method were discussed. Existing studies have made many achievements in topology improvement and control strategies, but the systematic theoretical system of power conversion technologies targeting the characteristics of aerospace pulsed power loads needs to be further improved, and challenges still exist in aspects such as dynamic response enhancement, multi-condition adaptation, and stability evaluation under strong time-varying conditions. The review can provide a comprehensive reference for the subsequent research and engineering applications of power conversion technologies for aerospace pulsed power loads.

As the construction of the international lunar research station enters the engineering implementation phase, efficient logistics transportation between lunar bases and mining areas becomes a core requirement to support lunar resource exploitation, while conventional wheeled lunar rovers face severe technical bottlenecks in low-gravity and soft lunar soil environments. To explore a novel lunar transportation scheme that breaks through the physical limits of wheeled systems, a maglev lunar rover concept based on Halbach permanent-magnet wheels was proposed; its technical feasibility and environmental adaptability were systematically demonstrated, and a brief review of the development history of related Chinese and international technologies was provided. First, the current dilemmas of wheeled lunar rover technology were sorted out from four aspects: development status, environmental constraints, failure modes, and performance bottlenecks. Second, the quadruple advantages of the adaptability of the maglev lunar rover to the lunar environment were explored from the perspectives of a non-adhesive driving mechanism, low-gravity load gain, in-situ resource pavement construction, and ultra-low-temperature electromagnetic enhancement, and a comparative analysis with existing lunar rover technologies was conducted. Finally, the research achievements of the team in electromagnetic modeling, dynamics analysis, stability control, and prototype verification of permanent-magnet wheel electrodynamic suspension vehicles were reviewed, and the key scientific issues and future research prospects for subsequent transplantation to the lunar environment were elucidated. The research results show that the performance bottleneck of conventional wheeled lunar rovers originates from the dependence of contact-type walking mechanisms on the physical environment of lunar soil, and institutions such as NASA have carried out exploration on the extraterrestrial application of maglev technology. The proposed permanent-magnet wheel maglev lunar rover features a simple structure and good environmental adaptability, and a closed-loop research of “electromagnetics-dynamics-control-prototype development” has been completed. As a conceptual scheme under lunar engineering conditions, key technologies including real environment simulation, high-latency autonomous communication regulation, and in-situ manufacturing of conductor-plate pavement need to be predominantly broken through in the future, so as to provide support for the construction of a reliable lunar transportation network.

Under the impetus of global climate governance and China’s carbon peaking and carbon neutrality goals, bridge engineering serves as a vital component of transportation infrastructure. Its green and low-carbon transformation is of great significance for reducing carbon emissions in the transportation sector and promoting sustainable development. The research progress, technical systems, and engineering applications of green and low-carbon technologies in bridge engineering were systematically reviewed. It covered the full life cycle of bridges, ranging from material, design, construction, operation, and maintenance to demolition, aiming to achieve coordinated emission reduction in the five aforementioned stages. The roles in reducing carbon emissions in the full life cycle, enhancing durability, and optimizing seismic performance in a coordinated manner were reviewed. On the material side, low-clinker/high-blended cementitious materials, recycled aggregates, and circular steel were used to reduce embodied carbon and enhance durability. On the design side, the coordinated optimization of life cycle cost and environmental impact was adopted to achieve carbon control decisions. On the construction side, direct emission reduction was achieved through prefabricated construction, efficient construction equipment, and optimized transportation, as well as organization. On the operation and maintenance side, building information modeling (BIM)/Internet of Things/artificial intelligence were utilized for condition perception and predictive maintenance to improve energy efficiency and reduce unnecessary maintenance activities. At the end of the life cycle, retirement-friendly design, hierarchical dismantling, as well as refined sorting, recycling, and remanufacturing technologies were employed to achieve the closed loop of material and final disposal reduction. Currently, there are still key issues such as the imperfection of multi-objective coordinated design methods, the lack of unified carbon emission evaluation standards and data systems, and the insufficient coordinated optimization of low-carbon and seismic performance. In the future, efforts should be made to strengthen the application of digital twin and intelligent optimization technologies, promote the research and development of low-carbon new materials, improve the green construction and intelligent operation and maintenance systems, and establish a material recycling and closed-loop management mechanism. The theory of coordinated design for low-carbon and seismic resilience should also be deepened to achieve the sustainable development goals of “safety, durability, low carbon, and resilience” in bridge engineering.

The transfer line balancing problem (TLBP) is widely applied in the production process of structural components in high-volume manufacturing. By reasonably allocating each processing element of parts to the machine tools in each workstation, efficient and balanced production of machining production lines is achieved. The research on TLBP is of great significance for improving product processing efficiency, quality, and output, as well as reducing comprehensive costs. TLBP has been studied for 27 years. However, due to the particularity of the problem and the complexity of the process, the models and solution methods for assembly line and disassembly line balancing problems cannot be directly applied to solving TLBP. As a result, the research on this problem has not received sufficient attention, and the existing literature is limited. Despite its long research history, there is still no systematic review in China that sorts out its past, current, and future development context. Therefore, a hierarchical-tuple composite representation method was adopted to systematically analyze the problem. The in-depth analysis and multi-dimensional display were conducted on the three key levels composing the problem framework: the level of problem structure types, the level of problem system configuration, and the level of problem decision support. A structured panoramic view of the field was provided, and the current research status was clarified. The research blind spots were refined, and possible future research directions were pointed out.

Deep integration of traction power supply systems in electrified railways and new energy is an important measure to implement the national “carbon peaking and carbon neutrality” goal and promote the green transformation of rail transit. In view of the demand for large-scale integration of new energy, combined with the national direct connection strategy of green electricity, a technical scheme for the direct connection of green electricity in traction power supply systems was proposed, and its power flow control strategy was studied. Firstly, different schemes for direct connection of green electricity were compared and analyzed, and it was clarified that through-feeding power supply with direct connection of green electricity was the optimal choice to achieve a win-win situation for power grids and railways. On this basis, by considering grid support strength and coupling degree, three architectures of grid-structured through-feeding power supply physical systems with direct connection of green electricity were constructed. Power supply zones were formed based on the demarcation of section posts, realizing the matching of generation, consumption, and energy storage within power supply zones. Secondly, control logic of green electricity devices and energy storage devices under grid-connected/islanded modes was established, and a regulation system driven by controller information flows to direct energy flows was constructed. Based on electrical quantity information of section posts, train operation states and traction load powers were identified, realizing real-time power balance and autonomous coordinated control of generation, storage, and consumption under grid-connected/islanded modes. Finally, taking an actual reconstructed line as an example, the effectiveness and economy of the scheme were verified. Research results indicate that the through-feeding power supply with direct connection of green electricity is conducive to large-scale integration of new energy along railways. Meanwhile, it can solve dual pain points of negative sequence of power grids and power interruption zones of neutral sections in traction networks, achieving the core goals of zero grid interference and zero operation interruption. Through segmented and partitioned power supply and autonomous coordinated control of power flows, energy autonomy, operation autonomy, and control autonomy of the system are realized within respective jurisdiction sections of traction substations and traction green electricity stations. Combined with line data for economic analysis, by adopting lithium iron phosphate battery energy storage devices and considering two operation scenarios of grid-connected and islanded modes, preliminary cost recovery periods are approximately 3.9 years and 7.2 years, respectively.

In the calculation of wheel–rail frictional temperature rise, temperature-dependent material properties and surface cracks induce material discontinuities and geometric discontinuities, respectively, which are difficult to handle using traditional analytical and finite element methods based on continuum mechanics. Therefore, based on the nonlocal peridynamic heat conduction theory, a two-dimensional analysis model for wheel–rail frictional temperature rise was established by using a moving heat source method to represent the frictional heat generation boundary in the wheel–rail contact region. First, under identical calculation parameters, the results of the established model were compared with those of the classical analytical method; subsequently, the effects of temperature-dependent material properties and adiabatic crack inclination angle on the rail frictional temperature rise were analyzed. The results indicate that the maximum rail surface temperatures obtained by the established model and the classical analytical method are 364.9 ℃ and 358.7 ℃, respectively, with a relative error of only 1.7%, which verifies the rationality and accuracy of the model. Without considering temperature-dependent material properties, the frictional temperature rise increases linearly with creepage; when considering temperature-dependent material properties, heat more easily accumulates near the rail surface, manifesting as increased surface temperature and decreased internal temperature, and this effect is more pronounced at higher creepage. Under the 15% creepage condition, the local internal temperature when considering temperature-dependent material properties is even lower than that under the 10% creepage condition without considering temperature-dependent material properties. Rail surface cracks significantly alter the local heat flow path and induce heat concentration. When the crack inclination angle is 30°, the peak temperature near the crack reaches 1 014.6 ℃, which is about three times the temperature at the same location in the crack-free condition. The research results provide a new numerical method for analyzing wheel–rail frictional temperature rise under complex conditions.

To investigate the damage mechanism of locomotive wheels during the braking of heavy-haul trains on long heavy downhill slopes with low adhesion, a wheel damage analysis model for heavy-haul locomotives considering longitudinal impulses was established focusing on the slave control locomotive of a 20,000-ton heavy-haul combined train. This model mainly included a longitudinal dynamics model of the 20,000-ton heavy-haul combined train, a vehicle-track spatial interaction model considering anti-slip control, and a wheel damage prediction model. Based on this model, the wheel damage distribution of the slave control locomotive under electric braking conditions was analyzed, and the effects of rail surface adhesion parameters and anti-slip control algorithm parameters on the wheel damage of the slave control locomotive were further explored. Simulation results indicate that: 1) Coupler forces lead to different wheel adhesion performance and sliding characteristics between Section A and Section B of the slave control locomotive when passing through low-adhesion areas, which in turn results in differences in the damage values among individual wheels of the slave control locomotive. Under normal rail adhesion conditions, the wheel damage is predominantly fatigue damage; with characteristic parameter matchings of different contact patches, the wheel damage gradually transforms from fatigue damage to wear damage. 2) Different anti-slip control thresholds exhibit varying effects on wheel damage under complex rail adhesion and braking conditions. When the electric braking force reduction ratio increases from 0.4 to 1.0, the maximum wheel damage value of each wheelset decreases by 88.9%. When the braking force reduction slope increases from 5 kN/s to 30 kN/s, the maximum decrease in the wheel damage value of each wheelset reaches 92.4%, and when the recovery slope decreases from 15 kN/s to 1 kN/s, the maximum decrease is 80.0%.

In complex mountainous areas, tunnel boring machine (TBM) construction often faces adverse geological conditions, such as fault fracture zones and large deformation of surrounding rock, which can easily cause jamming accidents. TBM jamming risk assessment involves multi-source heterogeneous information such as monitoring data and empirical rules. Existing studies mostly focus on the local modeling of assessment data or models and lack a systematic characterization of key assessment factors and their semantic associations, leading to a fragmented risk representation structure, which makes it difficult to support dynamic assessment and analysis under complex conditions. To address the above issues, a construction method of a knowledge graph for TBM jamming risk assessment based on multi-factor semantic associations was proposed. This method constructed an ontology structure by integrating five types of core factors: "assessment task, assessment data, assessment parameter, assessment model, and assessment indicator", achieving semantic association representation of the risk assessment process. Combined with a stepwise prompting strategy of a large language model, explicit and implicit knowledge was extracted from risk assessment literature, construction standards, and engineering cases, and the effectiveness of the prompting strategy was verified through ablation experiments. Furthermore, the Neo4j graph database was utilized to realize structured storage and management of the knowledge graph. The experimental results indicate that on the constructed annotated dataset, the precision, recall, and F1 score of knowledge extraction are 88.54%, 83.42%, and 85.67%, respectively, showing a significant improvement compared with using a single prompting strategy. The analysis of typical engineering cases shows that the proposed method can effectively organize factors related to jamming risk, exhibits a good performance in the completeness of risk information representation and semantic organization, and provides effective support for the structured representation and assessment of jamming risk.

To elucidate the evolution mechanism of longitudinal creep in ballasted continuous welded rail turnouts in extreme environments, based on time-series displacement data and resistance evolution characteristics, a refined analysis model for longitudinal creep of in-service ballasted continuous welded rail turnouts was established, with consideration of track resistance degradation effects. This model was developed through the analysis of resistance curve offsets, the decomposition of loading/unloading paths, and the search for the shortest equivalent path. The nonlinear longitudinal mechanical behavior of turnout rails and sleepers under extreme diurnal temperature cycles was elucidated for the first time, and the influence of updated fastening and ballast resistance parameters on residual stress and prediction results of cumulative creep displacement was assessed. The results indicate that cyclic loads with opposing directions and asymmetric amplitudes exacerbate longitudinal creep, with 15 extreme diurnal rail temperature cycles ranging from −50 ℃ to 10 ℃ resulting in a maximum residual rail deformation of 4.98 mm. Moreover, updating the parameters with a 20% degradation in both fastener and ballast resistance during the cycles causes the maximum residual rail deformation to increase to 5.48 mm, while the maximum residual stress of the nose rail grows by 20.24%.

In view of the diverse terrain types and highly variable spatial distributions of wind fields at bridge sites in complex mountainous terrain, engineering-identifiable terrain elements were taken as the main thread, and combined with research results such as field measurements and numerical simulations, a progressive classification framework based on “canyon cross-sectional morphology, canyon channel orientation, and special topography and geomorphology” was constructed. The influence laws of eight typical mountainous bridge site terrains (V-shaped/U-shaped/L-shaped cross-sections, Y-shaped/S-shaped orientations, and funnel-shaped contraction/reservoir-dam area/thermally driven wind) on relevant wind fields were clarified, and a comparative analysis was conducted combined with typical cases. The results indicate that the cross-sectional types determine lateral confinement and spatial non-uniformity; V-shaped deep-cut canyons have stronger near-surface boundary effects on sidewalls; the wind field in U-shaped deep and large canyons is more obviously constrained by channels; L-shaped canyons result in a non-uniform spatial distribution of the wind field under the influence of asymmetric terrains. Channel turning and confluence easily form acceleration and over-mountain transport; Y-shaped channels form a downstream acceleration zone accompanied by drastic variations in the vertical angle of attack; S-shaped sharp bends lead to wind speed enhancement after turning and reconstruction of the horizontal angle of attack. Among special geomorphologies, funnel-shaped contraction induces convergent strong winds; water level variation in reservoir-dam areas can cause large-angle-of-attack climbing in the near-dam area, while weakening the squeeze acceleration and smoothing the approach flow in the far field; thermally driven winds form clear diurnal cycle characteristics; the temperature difference over the underlying surface of snow mountains can enhance the convergence toward the valley floor. The constructed framework can support the rapid discrimination of wind environments and the identification of adverse working conditions at mountainous bridge sites and provides a reference for the wind-resistant design of mountainous bridges.

In order to meet the requirements for driving performance of railway bridges, it is necessary to control the smoothness of the bridge alignment. Based on the analysis of driving stability, the sensitive wavelength range of the vehicle body was determined. The amplitude of the completed bridge alignment within the range was adopted as the evaluation criterion. From the perspective of ensuring driving performance, by considering the self-adjustment capability of ballastless tracks and the relationship between the track surface alignment and the completed bridge alignment, the expression of the irregularity limit value for the completed bridge alignment was derived. A seven-span continuous steel truss girder bridge was taken as the research object. The irregularity amplitude of the completed bridge alignment within the sensitive wavelength range of the vehicle body was controlled according to the derived expression. A method for controlling the smoothness of the girder assembly alignment based on the target alignment of the completed bridge was proposed by combining Akima spline curve with the girder assembly curve. The research results show that the sensitive wavelength of the vehicle body is less than 200 m at driving speeds of trains ranging from 250 km/h to 350 km/h. If a seven-span continuous steel truss girder bridge is taken as an example, the irregularity limit values for the completed bridge alignment within the sensitive wavelength ranges of the vehicle body corresponding to speeds of 350 km/h, 300 km/h, and 250 km/h are 24, 26 mm, and 29 mm, respectively. The irregularity amplitude within a 0–200 m wavelength range of the girder assembly alignment can be evaluated and controlled through the proposed method for controlling the smoothness of the girder assembly alignment.

To clarify the occurrence characteristics and ecological risks of pollutants (e.g., heavy metals and organophosphate esters) in the construction wastewater of drill-and-blast tunnels, two tunnels under construction on the Qinghai-Tibet Plateau were taken as the research objects, and the water quality parameters, pollution characteristics, sources, and environmental risks of the construction wastewater and surrounding water bodies were analyzed. The research results indicate that the construction wastewater exhibits the characteristics of high turbidity (average of 43.7–100 NTU) and alkalinity (pH reaches 8.2–12.0) and contains high concentrations of petroleum pollutants (average concentration of 15.5–22.1 mg/L), which are mainly derived from pollutions such as mechanical lubricants; the metal elements are mainly Fe and Al, and the anions are mainly Cl⁻ and SO₄²⁻, whose main sources are the use of accelerators/coagulants and the dissolution and release of minerals; the total concentration range of OPFRs is 14.0–6 060 ng/L, and the main components include tributyl phosphate and tris (2-chloroethyl) phosphate, which may be sourced from construction materials and lubricants; the environmental risk assessment shows that the maximum value of the risk quotient of 2-ethylhexyl diphenyl phosphate in the tunnel construction wastewater and surface water reaches 0.68–1.83, indicating that it has moderate to high ecological risks, and the accumulation of tris (2-chloroisopropyl) phosphate and trioctyl phosphate in the tunnel construction wastewater is worthy of attention. The migration and transformation patterns of composite pollution during the construction process of high-altitude tunnels are revealed, providing a scientific basis for the optimization of green construction technologies and the precise control of pollutants.

To address the low accuracy of numerical simulation in traditional friction models during the prediction of the forming process of ultra-high strength steel, the influence of sliding speed and normal load on the friction behavior of ultra-high strength steel CP780 was investigated using a self-developed stamping friction testing machine for ultra-high strength steel. A dynamic friction coefficient model related to sliding speed and normal load for ultra-high strength steel stamping and forming was established, and the proposed model was verified by integrating U-bending experiments with numerical simulations. Research findings indicate that the friction coefficient of CP780 sheet materials increases with increasing sliding speed and decreases with increasing load. Under conditions of low load and low speed, the wear mechanism of CP780 sheet materials is primarily dominated by ploughing effects; under conditions of high speed and high load, the wear mechanism involves both ploughing effects and partial adhesive effects. A comparison of rebound test values from U-bending experiments with numerical simulation results reveals that the error in the rebound angle α predicted by the dynamic friction model is 1.509%, and the error in the predicted rebound angle β is 0.348%. In contrast, the error in the rebound angle α predicted by the traditional constant friction coefficient model is as high as 12.483%, and the error in the predicted rebound angle β is as high as 4.994%. The dynamic friction model developed in this study demonstrates greater precision in predicting rebound angles and significantly enhances the accuracy of numerical simulations for formed parts.

To address the difficulty of balancing fuel economy and durability, an energy management strategy (SAC-V) for fuel cell hybrid electric buses was proposed by combining a bidirectional long short-term memory neural network (BiLSTMNN) with a soft actor-critic (SAC) algorithm. First, BiLSTMNN was applied to achieve short-term vehicle speed prediction. Second, the predicted speed and real-time vehicle states were jointly used as inputs to the SAC reinforcement learning agent, and constraint terms including hydrogen consumption, power battery state of charge deviation, and fuel cell degradation were introduced into the reward function to achieve dynamic coordinated optimization of vehicle fuel economy and powertrain durability. Finally, the proposed strategy was validated through offline simulation and hardware-in-the-loop tests. The research results show that, compared with the conventional SAC method and the deep deterministic policy gradient method, the proposed SAC-V strategy reduces equivalent hydrogen consumption, fuel cell power fluctuation, and fuel cell degradation rate by 3.46%, 35.56%, and 3.67%, respectively, exhibiting better overall performance and favorable real-time application potential in practical engineering.