Shuai Ma

Vibration comfort is one of the most influencing factors on the service quality of high-speed railways, and it is meaningful to relate vehicle vibrational responses to track geometry activation for the purpose of locating adverse track states from comfort point of view. Although a wide variety of mechanical and data-driven models have been put forward for this task, these models are unsatisfactory in computational efficiency or accuracy. In this circumstance, a deep learning model, integrating convolutional neural network, long short-term memory, and multilayer perceptron, is proposed to automatically learn the complex relationship between track geometries and vehicle vibrations. The trained model can predict local vehicle-body accelerations and sectional comfort indices simultaneously by taking track geometries and running velocity as input. According to the evaluation test on the inspection data of a high-speed railway in China, the predictions are in good consistency with the real values, with outperforming efficiency in time cost compared with the running velocity of high-speed vehicles. Moreover, the predictions made by the model can in turn be regarded as assessment indices of track geometric state, for revealing adverse track states or potential track defects. The predicted comfort index is superior to traditional track quality index in revealing vibrational comfort of the running vehicle, promising in acting as a supplement, or even a replacement, to the latter.

Track geometry inspection data have long endured positional errors and shifts, hindering the quality evaluation, deterioration analysis and maintenance work of railway track. Besides, vehicle response inspection data are also typically misaligned or shifted from each other, but it tends to be ignored by people. In light of this situation, a position calibration method on the basis of a simple similarity metric, i.e. correlation coefficient, is proposed to tackle the positional errors and shifts among multisource track inspection data concurrently. For the convenience of practical implementation, a scheme containing three parts is designed to fulfil the purpose and a two-step procedure is adopted by all the three parts. According to the performance test on a dataset containing five inspection runs of a high-speed railway in China, the positional error between inspection data and actual locations can be greatly diminished, and the data inspected at different times and from different sources are well aligned with each other, with a time consumption of 0.66 s on average for a 1 km segment.

High-speed railway administrations are particularly concerned about safety and comfort issues, which are sometimes threatened by the differential deformation of substructures. Existing deformation-monitoring techniques are impractical for covering the whole range of a railway line at acceptable costs. Fortunately, the information about differential substructure deformation is contained in the dynamic inspection data of longitudinal level from comprehensive inspections trains. In order to detect potential differential deformations, an identification method, combining digital filtering, a convolutional neural network and infrastructure base information, is proposed. In this method, a low-pass filter is designed to remove short-waveband components of the longitudinal level. Then, a one-dimensional convolutional neural network is constructed to serve as a feature extractor from local longitudinal-level waveforms, and a binary classifier of potential differential deformations in place of the visual judgement of humans with profound expertise. Finally, the infrastructure base information is utilized to further classify the differential deformations into several types, according to the positional distribution of the substructures. The inspection data of four typical high-speed railways are selected to train and test the method. The results show that the convolutional neural network can identify differential substructure-deformations, with the precision, recall, accuracy and F1 score all exceeding 98% on the test data. In addition, four types of deformation can be further classified with the support of infrastructure base information. The proposed method can be used for directly locating adverse substructure deformations, and is also becoming a promising addition to existing deformation monitoring methods.

The problem of excessive wear of pantograph strips frequently occurs on China’s Z City Metro Line 1. After on-site investigation and analysis by the metro operating company, it was speculated that the problem was related to abnormal track irregularities. Therefore, taking Z City Metro Line 1 as the main research object, the measured track irregularity of the whole line was analyzed and compared with other typical track spectra, and a track-vehicle-pantograph-catenary coupling dynamics model was established to analyze the relationship between the pantograph-catenary dynamic characteristics and the track irregularity. The two frequency ranges of the track irregularity that have a significant impact on the pantograph-catenary contact were found. Finally, after numerical calculation and analysis, it is recommended to focus on the irregularities with the two wavelength ranges of 9~16 m and 3~4 m in the maintenance.

At present, all systems of armored vehicles are function intensive and software intensive systems, and the physical connection between the components is difficult to be accurately represented by models. Moreover, the faults are characterized by concealment, correlation and coupling, which makes fault diagnosis difficult. In this paper, the dependency model based on typical control system is established, and the test time, test cost, test reliability and other influencing factors of fault diagnosis are comprehensively considered. The optimal diagnosis test sequence is quickly generated, which can realize the rapid fault location, so as to improve the accuracy and comprehensiveness of the test, and reduce the dependence on the experience of designers, Give full play to the combat effectiveness of armored vehicles and weaponry.

During the long-term service of slab track, various external factors (such as complicated temperature) can result in a series of slab damages. Among them, slab arching changes the structural mechanical properties, deteriorates the track geometry conditions, and even threatens the operation of trains. Therefore, it is necessary to detect slab arching accurately to achieve effective maintenance. However, the current damage detection methods cannot satisfy high accuracy and low cost simultaneously, making it difficult to achieve large-scale and efficient arching detection. To this end, this paper proposed a vision-based arching detection method using track geometry data. The main works include: (1) data nonlinear deviation correction and arching characteristics analysis; (2) data conversion and augmentation; (3) design and experiments of convolutional neural network- based detection model. The results show that the proposed method can detect arching damages effectively, and the F₁-score reaches 98.4%. By balancing the sample size of each pattern, the performance can be further improved. Moreover, the method outperforms the plain deep learning network. In practice, the proposed method can be employed to detect slab arching and help to make maintenance plans. The method can also be applied to the data-based detection of other structural damages and has broad prospects.

Track quality evaluation is fundamental for track maintenance. Around the world, track geometry standards are established to evaluate track quality. However, these standards may not be capable of detecting some abnormal track geometry conditions that can cause considerable vehicle-body vibration. And people gradually realized that track quality evaluation should be based not only on track geometry but also on vehicle performance. Vehicle-body vibration prediction is beneficial for locating potential track geometry defects, and the predicted accelerations can be used as an auxiliary index for assessing track quality. For this purpose, this paper gives a method to predict vehicle-body vibration based on deep learning, which represents one of the newest areas in artificial intelligence. By integrating convolutional neural network (CNN) and long short-term memory (LSTM), a CNN-LSTM model is proposed to make accurate and point-wise prediction. To achieve the optimal performance and explore the internal mechanism of the model, structural configurations and inner states are extensively studied. CNN-LSTM can take advantage of the powerful feature extraction capacity of CNN and LSTM, and outperforms the fully-connected neural network and the plain LSTM on the experimental data of a high-speed railway. In detail, CNN-LSTM has superior performance in predicting vertical vehicle-body vibration below 10 Hz and lateral vehicle-body vibration below 1 Hz. Moreover, analysis shows that the predicted vehicle-body acceleration can act as a performance-based evaluation index of track quality.