Railway turnouts are critical but weak devices that branch one track into two or more. To ensure the safety of train passage, the insufficient displacement (ID) of the switch rail must be controlled within the deviation required by the railway engineering. A reasonable tractive layout is the key to maintaining the ID of the switch rail at a relatively low level after its operation. Existing research generally formulates the design problem as a non-convex problem and cannot guarantee global convergence. To tackle this issue, this paper proposes a sparse learning-based design method for the tractive layout of the switch rail. Firstly, by defining the tractive force vector as the indicator variable, we transform the original combinatorial optimization problem into a continuous one and formulate a constrained convex problem by combining the$\boldsymbol {L}_{\mathbf {1}}$norm and the weighted squared ID. Then, an ADMM-based algorithm is developed to efficiently solve the force vector, thereby adaptively selecting the tractive layout. Finally, we validate the effectiveness of the proposed method using two typical turnouts in railway industry.

The rail surface defects directly impact the safety and efficiency of heavy-haul train operations. Timely assessment of these defects is crucial for informed maintenance decisions, with precise defect detection at its core. In recent years, the accumulation of extensive rail inspection images has led to the application of numerous computer vision-based methods for pixel-level detection of rail surface defects. However, given the constraint of a limited number of labeled defect samples, ensuring the generalization and robustness of existing methods remains challenging, particularly across varying track conditions and complex heavy-haul scenarios. Thus, this paper introduces a Segment-Anything-Model (SAM)-enhanced method for the detection of rail surface defects. First, a shadow-detection-based algorithm is developed to extract the rail regions and mitigate background interference. Then a student-teacher-Simi-network (S-T-Simi)-based unsupervised method is designed to generate prompt information for SAM. Utilizing this prompt information, we develop a task-specified SAM for precise rail defect detection. Finally, comprehensive validation is performed using inspection data collected from diverse heavy-haul tracks. Experimental results indicate that the proposed method achieves highly accurate segmentation of rail defects.

The diagnosis of railway system faults is significant for its comfort, efficiency, and safety. The rail profile faults are the most direct impact factors when considering the health conditions of rails. This paper puts forward rail fault diagnosis from two perspectives: quantifying the level of complexity and chaos of different profiles, and measuring the level of correlation between different profiles, which correspond to the newly proposed refined dispersion entropy (RDE) method and the correlation plane method, respectively. The RDE uses weighted-dispersion patterns to extract accurate time domain features from rail profile data, and the correlation plane can characterize nonlinear and non-monotonic relationships between analyzing subjects, which are the main contributions of this study. Experimental results with simulated and reality-based data show that the proposed methods can identify faulty profile data and discriminate different types of profile faults more effectively when compared with existing methods.

With advancements in sensor technology, high dimensional signals such as functional curves and images are typically collected from multiple sensors to characterize the degradation of a system. Data fusion methods are employed to integrate multisensor signals generated from the system into a scalar health index (HI) to understand the degradation status of the system. This paper develops sparse group LASSO-principal component analysis (SGL-PCA), a method that constructs HIs for image and profile data. First, we remove the smooth background from each sensor signal. Then, we solve the degradation patterns and the degradation paths through a rank-one matrix approximation problem, with the consideration of the sparsity of the measurements related to the degradation process and the monotonicity of the degradation paths. Results from a simulation study and a case study illustrate that the HI constructed by the proposed method outperforms the benchmark methods in identifying the measurements subject to the degradation process and predicting the remaining useful life of the system. Note to Practitioners—In practice, sensors generating multiple high-dimensional curves and images are often installed in systems to characterize their degradation status. Compared with scalar sensor signals, the information that associates with the degradation process often appears in sparse regions from the sensor signals. Therefore, identifying the degradation information accurately is important in the health index (HI) construction for degradation modeling and prognostic analysis. This article proposes a method that simultaneously selects the degradation information and estimates the optimal weights for integrating multi-sensor signals in constructing the HI. The proposed method is applicable in the case where the systems degrade under a single failure mode, and multiple sensors are used to monitor the degradation processes. Practitioners can implement our method to predict the remaining useful lives of in-service systems through three steps: (1) estimate the backgrounds of each sensor and derive data-fusion model using a historical dataset; (2) construct the HIs of in-service systems; (3) predict the remaining useful lives of these systems based on the developed HIs.

Railway turnout is the key infrastructure for trains to change their routes. In order to ensure the smoothness and safety of the train’s passing through turnouts, the insufficient displacement (ID) of switch rails after their conversions must be controlled within a permitted range. During design stage, it is quite important for the reduction of the ID to reasonably arrange the tractive points. In the existing literature, a feasible tractive layout is commonly suggested through the manual analysis using the finite element model of the switch rail. However, as the design space may explode for long rails with multiple tractive points, it is time-consuming for such labor-intensive methods to search a feasible tractive layout, and the result may be non-optimal. Therefore, it is necessary to develop an efficient method to arrange the tractive points optimally in order to minimize the ID. To this end, we propose a physics-informed optimization method for the design of the tractive layout. First, a tailored direct stiffness method is introduced to accurately estimate the ID given any tractive layout and frictions. On this basis, we establish an optimization model for the selection of the tractive locations with the objective of minimizing the expectation of the ID. To address the sparsity issue of the decision variable, an Encoding Rule with a hierarchical indexing method is proposed to improve the efficiency of genetic algorithm. Next, the number of tractive points is determined. Finally, several sets of experiments are conducted to demonstrate the effectiveness of the proposed method, which decreases the ID by 25.61% and 12.7% in terms of the maximal and mean values for the case with the switch of length 44.1m.

Heterogeneous but complementary sources of data provide an unprecedented opportunity for developing accurate statistical models of systems. Although the existing methods have shown promising results, they are mostly applicable to situations where the system output is measured in its complete form. In reality, however, it may not be feasible to obtain the complete output measurement of a system, which results in observations that contain missing values. This article introduces a general framework that integrates tensor regression with tensor completion and proposes an efficient optimization framework that alternates between two steps for parameter estimation. Through multiple simulations and a case study, we evaluate the performance of the proposed method. The results indicate the superiority of the proposed method in comparison to a benchmark. Note to Practitioners—The proposed method aims to obtain an accurate estimation of the regression model when certain entries of the response are inaccessible. By considering both the information from multiple inputs and the structure of the response, our proposed method can achieve more accurate estimation of the output tensor. In order to apply the proposed method in practice, two assumptions should hold. First, the response tensor should be low-rank, meaning that fewer variation patterns should exist in the response than its dimensions. Second, the relationship between the input tensors and the response should be linear or approximately linear. The presented method in this article uses tensor decomposition techniques to exploit the correlation structures of the high-dimensional data and prevent overfitting. Another benefit of our integrated framework is that the rank of the response tensor converges automatically, which can be used directly in the parameter estimation.

Degradation modeling is a critical and challenging problem as it serves as the basis for system prognostics and evolution mechanism analysis. In practice, multiple sensors are used to monitor the status of a system. Thus, multisensor data fusion techniques have been proposed to capture comprehensive information for prognostic modeling and analysis, which aims at developing a composite health index (HI) through the fusion of multiple sensor signals. In the literature, most existing methods use a linear data-fusion model for integration of multisensor data to construct the HI, which is insufficient to model nonlinear relations between sensing signals and HI in a complicated system. This article proposes a novel data fusion method based on deep learning for HI construction for prognostic analysis. A pair of adversarial networks is proposed to enable the training procedure of neural networks. To guarantee the stability of the algorithm, we propose a root mean square propagation (i.e., RMSprop)-based sampling algorithm to estimate model parameters. A set of simulation studies and a case study on a set of degradation signals of aircraft engines are conducted. The results demonstrate that the proposed method has a significant improvement on remaining useful life prediction compared to existing data fusion methods.

A vast amount of text data is recorded in the forms of repair verbatim in railway maintenance sectors. Efficient text mining of such maintenance data plays an important role in detecting anomalies and improving fault diagnosis efficiency. However, unstructured verbatim, high-dimensional data, and imbalanced fault class distribution pose challenges for feature selections and fault diagnosis. We propose a bilevel feature extraction-based text mining that integrates features extracted at both syntax and semantic levels with the aim to improve the fault classification performance. We first perform an improved X2statistics-based feature selection at the syntax level to overcome the learning difficulty caused by an imbalanced data set. Then, we perform a prior latent Dirichlet allocation-based feature selection at the semantic level to reduce the data set into a low-dimensional topic space. Finally, we fuse fault features derived from both syntax and semantic levels via serial fusion. The proposed method uses fault features at different levels and enhances the precision of fault diagnosis for all fault classes, particularly minority ones. Its performance has been validated by using a railway maintenance data set collected from 2008 to 2014 by a railway corporation. It outperforms traditional approaches.

In automated manufacturing systems (AMSs), deadlock problems must be well solved. Many deadlock control policies, which are based on siphons or Resource-Transition Circuits (RTCs) of Petri net models of AMSs, have been proposed. To obtain a live Petri net controller of small size, this paper proposes for the first time the concept of transition covers in Petri net models. A transition cover is a set of Maximal Perfect RTCs (MPCs), and the transition set of its MPCs can cover the set of transitions of all MPCs. By adding a control place with the proper control variable to each MPC in an effective transition cover to make sure that it is not saturated, it is proved that deadlocks can be prevented, whereas the control variables can be obtained by linear integer programming. Since the number of MPCs in an effective transition cover is less than twice that of transition vertices, the obtained controller is of small size. The effectiveness of a transition cover is checked, and ineffective transition covers can be transformed into effective ones. Some examples are used to illustrate the proposed methods and show the advantage over the previous ones.

Deadlock-free control and scheduling are vital for optimizing the performance of automated manufacturing systems (AMSs) with shared resources and route flexibility. Based on the Petri net models of AMSs, this paper embeds the optimal deadlock avoidance policy into the genetic algorithm and develops a novel deadlock-free genetic scheduling algorithm for AMSs. A possible solution of the scheduling problem is coded as a chromosome representation that is a permutation with repetition of parts. By using the one-step look-ahead method in the optimal deadlock control policy, the feasibility of a chromosome is checked, and infeasible chromosomes are amended into feasible ones, which can be easily decoded into a feasible deadlock-free schedule. The chromosome representation and polynomial complexity of checking and amending procedures together support the cooperative aspect of genetic search for scheduling problems strongly.

The resource-transition circuit ( RTC) and siphon are two different structural objects of Petri nets and used to develop deadlock control policies for automated manufacturing systems. They are related to the liveness property of Petri net models and thus used to characterize and avoid deadlocks. Based on them, there are two kinds of methods for developing deadlock controllers. Such methods rely on the computation of all maximal perfect RTCs and strict minimal siphons (SMSs), respectively. This paper concentrates on a class of Petri nets called a system of simple sequential processes with resources, establishes the relation between two kinds of control methods, and identifies maximal perfect RTCs and SMSs. A graph-based technique is used to find all elementary RTC structures. They are then used to derive all RTCs. Next, an iterative method is developed to recursively construct all maximal perfect RTCs from elementary ones. Finally, a one-to-one correspondence between SMSs and maximal perfect RTCs and, hence, an equivalence between two deadlock control methods are established.