Abstract:In order to investigate the influence of gauge rods on the dynamic performance in tight curves of heavy haul railway, the wheel-rail dynamic interaction and wheel-rail wear when the locomotive passed the R300 m curve at a running speed of 70 km/h were analyzed on the basis of vehicle-track coupling dynamic theory. The influence of running speed and curve radius on gauge dynamic expansion, wear number, and gauge rods were analyzed. Furthermore, the influence of the spacing of gauge rods on the lateral stability of the track was studied. The simulation results indicate that the gauge rods can stabilize the gauge and reduce the turning angle of the rail at the outside curve. Compared with the curve without gauge rods, the contact point of the inner rail of the curve with gauge rods is closer to the inner side of the curve. The wear number and the dynamic gauge expansion when the locomotive negotiates a tight curve will increase with the decrease of curve radius and the increase of running speed. Increasing the arrangement density of gauge rods can effectively enhance the ability to stabilize the gauge. The dynamic gauge expansion will reduce by 36.3% when the spacing of gauge rods is reduced from 4 to 3 rail spans.

Abstract:Rockburst in deep tunnels is a hazard during the underground engineering construction. Accurate early warning of rockburst can protect the lives and properties of engineering personnel. The intelligent technologies such as machine learning （ML） have brought new ideas and methods for rockburst early warning, which has improved the accuracy, timeliness and intelligence for early warning of rockburst. A systematic study on the current application of ML in microseismic （MS） monitoring and early warning of rockbursts in deep tunnels was carried out. Firstly, ML algorithms in the MS monitoring, evaluation and early warning of rockbursts were summarized. The characteristic advantages of the various types of ML algorithms were analyzed. Then, the indicator system for MS monitoring and early warning of rockburst was discussed. The applications of MS monitoring and early warning of rockburst based on different ML methods and their effects were analyzed. The results show that neural network （NN） is one of the most popular algorithms for rockburst warning, the MS event（N）, MS energy（E）, MS apparent volume （V） and its variants are the most frequently used MS parameters, and most of the rockburst warning parameters are between 3~7 in number. Rockburst intensity is the research hotspot of rockburst warning, and the warning accuracy based on most ML methods can reach 80%, which indicates that the ML method has good application effects and development prospects. Finally, prospects were made for the development direction of ML in MS monitoring and early warning for the rockburst in deep tunnels, i.e. advanced ML algorithms, the accuracy and comprehensiveness of the early warning indicator system, the richness of the sample, the time warning of the rockburst occurrence, and the capability of data processing to be further investigated in depth.

Abstract:Based on the composition and characteristics of pavement defect detection system, this paper briefly reviews the development process of the pavement defect detection system. On this basis, it analyzes the status quo of typical pavement defect detection systems at home and abroad, including the intelligent detection system of heavy pavement condition and lightweight pavement quality detection system, and describes the performance and some parameters of the detection system.Then emphatically introduces the evolution process of pavement defect detection technology and methods from traditional image processing to the intelligent pavement defect detection method based on machine learning and deep learning theory is explored. And, the research progress of intelligent pavement defect detection methods based on deep learning technology at home and abroad is comprehensively introduced, including pavement defect detection methods based on regional convolutional neural network, single multi-frame detector, YOLO target detection, Transformer detection model, etc.Finally, the development trend and application prospect of intelligent detection system for pavement defects are discussed from the aspects of multi-mode information fusion, dual lightweight design and robust intelligent algorithm.

Abstract:Aiming at the problem that it is difficult to extract the fault features of railway locomotive bearings in a real complex environment, which leads to the difficulty of fault diagnosis, an improved multiscale symbolic dynamic entropy（IMSDE） fault diagnosis method is proposed. Firstly, the MSDE is improved by utilizing neighborhood slip averaging, which overcomes the defects of entropy deviation caused by traditional coarse -graining. Then, IMSDE is used to fully extract the key fault features of vibration signals at different scales. Finally, the identification of different fault types and degrees of railway bearings is achieved by combining with an extreme learning machine（ELM）. On this basis, three separate sets of tests were analyzed. The results show that the method has an accurate fault identification effect for both artificially constructed bearing faults and bearing faults generated by engineering reality, and the fault identification rate is higher compared with the other four methods, which verifies that the method has a certain value of practical application in engineering.

Abstract:Aiming at the problem of serious interference and low detection accuracy of insulator pictures collected in UAV patrol inspection，the optimization is carried out based on YOLOv5s algorithm，and the simulation research of insulator fault detection is carried out based on the improved YOLOv5s algorithm. The original algorithm is improved by adding CBAM attention module to the neck network，using K-means clustering to recalculate the size of a priori frame，and using MetaAconC as the activation function. The experimental results are analyzed based on Python. The experimental results show that the advantage of the proposed scheme is that the average accuracy of the algorithm mAP reaches 96.7%，which is 3.3% higher than the original YOLOv5s model; In addition，the weight file size of the algorithm training in this scheme is only 15.1 M，which is only 0.1 M larger than the original YOLOv5s. With the lightweight feature，the proposed scheme has a good prospect in the deployment of intelligent patrol work.

Abstract:In order to identify the traffic operation law of urban arterial road and support basis for traffic management departments to formulate relevant traffic demand management policies, a vehicle travel group identification model of urban arterial road based on combined model was proposed. In this study, a travel characteristic indicator system was constructed from dimensions of travel intensity, travel time, travel habits for comprehensively describing the travel behavior based on the traffic bayonet data of Qingdao Jiaozhou Bay Tunnel. The redundant indicator was eliminated based on the correlation analysis to avoid the impact on identification research. For the mixed attribute travel characteristic indicator data, the improved K-prototypes algorithm was used to effectively classify the vehicle travel groups, and combined with GBDT, the identification model based on improved K-prototypes and GBDT was established. By randomly selecting 10 000 samples to conduct identification research, the result shows that there are 5 vehicle travel groups for the road in this research, including high-frequency commuter groups, low-frequency commuter groups, operating groups, frequency stable groups, and ordinary groups. For the 5 vehicle travel groups, the average identification accuracy rate exceeds 97.75%, and the highest identification accuracy rate can reach 99.47%.

Abstract:The operation environment of high-speed trains is complex and changeable. The existing target curve of given operation speed mainly considers the safety and punctuality of train operation, which is difficult to improve other operation performance of trains. In order to meet the increasing traffic demand of high-speed trains and improve the running performance of trains, this paper proposes an improved multi-objective speed optimization method for safety, energy saving, punctuality and comfort, considering the optimal adhesion between wheels and rails. First of all, on the premise of meeting the restriction of interval speed limit and train dynamics model, four evaluation indexes of safety, energy saving, punctuality and comfort are established to form a multi-objective optimization model of high-speed train operation process; Secondly, the influence of adhesion between wheel and rail is considered in the energy-saving model, and the traction/braking force is optimized to keep it within the optimal adhesion range, so as to save operating energy consumption; Finally, optimization of multi-objective operating speed curve uses reference point based non dominated sorting optimization algorithm（NSGA-Ⅲ）. The simulation results of real lines show that the optimization effect considering wheel rail adhesion is significantly improved, especially in energy saving; Compared with GA and NSGA-Ⅱ, the NSGA-Ⅲ algorithm has better convergence effect and convergence speed.

Abstract:In order to effectively judge the geometric state of track and adapt to the maintenance state of highspeed railway, the optimization of track irregularity maintenance strategy is studied. The Markov decision process is selected, the model parameters such as track state level, maintenance action space and maintenance action cost are set, and the value iteration algorithm is used to solve the problem, so as to realize the effective formulation of high-speed railway line maintenance plan. Taking a ballasted high-speed railway line in East China as an example, combined with the characteristics of the track unit at typical deterioration speed, the maintenance decision optimization process of the Markov decision model is analyzed and the effect is verified. At the same time, the optimal maintenance decision of the track unit section at each decision time is explored. Monte Carlo stochastic simulation is used to simulate the total maintenance cost in the planning cycle and compare it with the actual maintenance cost. The results show that the track irregularity maintenance decision based on Markov decision process can fully consider the heterogeneity of track unit section irregularity deterioration, scientifically arrange maintenance activities according to the actual state and deterioration law of track unit section, improve the spatial resolution of maintenance operation, and the optimization effect of maintenance decision in the planning cycle is remarkable, which reduces the maintenance cost while ensuring the high smoothness of the line, and has a guiding role in the maintenance and repair of railway track.

Abstract:The study of train speed profiles plays a crucial role in optimizing the train operation process for urban rail train. To achieve better optimization results of train speed profiles, aiming at the three goals of train running punctuality, running energy consumption and passenger comfort, an approach based on an improved multi-objective differential evolution algorithm is proposed. Firstly, a multi-objective optimization model for urban rail transit trains is established based on the train operation process. Then, by adopting an elite mirror initialization strategy, ntroducing parameter adaptation and multi-mutation strategies, the performance of the multi-objective differential evolution (MODE) algorithm is improved, and by comparing with the IGD values obtained by the other 6 comparison algorithms on the ZDT series test functions, the superiority of the proposed algorithm is verified. Finally, combined with the real line data of Nanchang Metro Line 1, the simulation results show that the improved MODE algorithm has certain advantages in comprehensive performance compared with the comparison algorithm, and has strong practicability in train energy-saving optimization problems.

Abstract:To study reasonable seismic design parameters of track structures, based on the theory of structural dynamics, a track dynamics analysis model considering seismic excitation sources was established. The dynamic response of CRTSⅢ slab ballastless track caused by seismic excitation was calculated. The influence of track structure parameters on dynamic response variables were studied. The findings show that the frequency of the typical earthquake waves is mainly in the range of 0~10 Hz. The three-dimensional natural frequencies of the track structure are greater than 10 Hz. The track displacement under earthquake may exceed the standard limit requirements, and the composite track slab can be connected to the base to strengthen the seismic resistance of the track structure. When the stiffness of the fasteners or the elastic modulus of the base increases, the dynamic response indicators basically increase accordingly. When reasonable matching of track parameters is ensured, appropriately reducing the stiffness of the fasteners and the elastic modulus of the base is beneficial for structural seismic design. When the stiffness of the fastener or the elastic modulus of the base changes, the lateral displacement of the track slab, the longitudinal stress of the base, and the lateral acceleration of the rail are significantly changed, which should be paid attention to during the calculation. The results can provide reference for seismic design, evaluation, and reinforcement measures of track structures.

Abstract:Railway bridges are a major project in railway construction, and the construction quality has a significant impact on the overall engineering quality. With the development of intelligent technology, applying technologies such as BIM, AI, VR, and GIS to railway bridge construction can improve construction level and efficiency. By establishing a parameterized BIM model, dynamic simulation of the construction process can be achieved, deepening the application of BIM in quality, and achieving intelligent and information based construction. The integration of these new concepts and technologies will bring more possibilities for railway bridge construction.

Abstract:In order to realize the intelligent analysis of longitudinal force and deformation of CRTSⅡ slab ballastless track CWR on bridge, this study considers the interaction between bridge structure, track slab and rail, and establishes a refined finite element model of CRTSⅡ slab ballastless track CWR on multi-span simply supported beam and long-span continuous beam bridge by using finite element method. The re-development of ANSYS is carried out by using C# language. An intelligent analysis system of longitudinal force is developed, which integrates parameter inputting, finite element modeling, loading application, automatic calculation, data extraction and intelligent data processing. The analysis system′s universality and reliability are confirmed through a comparison with previous research findings. The analysis system can provide reference for the design, operation and maintenance of typeⅡ slab ballastless track CWR on bridges.

Abstract:Railway subgrade design and construction is an important part of railway infrastructure construction, which plays an important role in ensuring the safety and efficiency of railway operations. However, there are a series of problems and challenges, such as low design efficiency, difficult construction quality monitoring, poor design and construction coordination, which cannot adapt to the needs of efficient, accurate and sustainable railway subgrade construction. The rapid development of BIM technology, as a new design and management tool, has become an important auxiliary technology for modernized railway subgrade design and construction, which brings great opportunities and challenges for the railway subgrade. This paper studies the application status of railway subgrade design and construction based on BIM technology and discusses the challenges and development trends. It found that BIM technology in railway subgrade design and construction has the advantages of improving design accuracy, reducing errors and conflicts, and optimizing construction management. However, there are a lot of challenges in technology, organization and management, and legal aspects. In the future, BIM technology in railway subgrade design and construction will develop in the direction of intelligence, collaboration and digitalization.

Abstract:Aiming at the problems of low segmentation accuracy and low detection efficiency of split pin defect detection algorithms based on semantic segmentation, this paper proposes an improved method of split pin defect detection for catenary based on DeepLabv3+. Firstly, the MobileNetv2 network is pruned, and the MobileNetv2 network is replaced with Xception for feature extraction, which can greatly reduce the consumption of computing resources and improve the detection efficiency. Then, CBAM(Convolutional Block Attention Module) is integrated into ASPP(Atrous Spatial Pyramid Pooling) module, and CBAM is introduced to process shallow features of input Decoder network, enhance the perception of split pin edge features, and improve the accuracy of model semantic segmentation. In order to alleviate the negative impact caused by the imbalance between split pin region and background region and improve split pin segmentation accuracy, CEDice Loss is used as the Loss function in this paper, combining the advantages of Cross -Entropy Loss and Dice Loss. Finally, according to the principle of split pin defect discrimination formulated in this paper, the state recognition of split pin is carried out according to the color and shape information of image segmentation. The experimental results show that compared with the original DeepLabv3+ model, the MPA and MIOU of the improved DeepLabv3+ model are improved by 3.54% and 3.42%, respectively, and the testing time is reduced by 14.41 ms per image, and the model parameters is reduced by 88.61%. In terms of defect identification, the accuracy of the method for missing, loose and normal split pins is 100%, 98.1% and 99.5%, respectively, which can quickly and effectively identify split pin defects.