Deep-neural network-based fault diagnosis methods have been widely used according to the state of the art. However, a few of them consider the prior knowledge of the system of interest, which is beneficial for fault diagnosis. To this end, a new fault diagnosis method based on the graph convolutional network (GCN) using a hybrid of the available measurement and the prior knowledge is proposed. Specifically, this method first uses the structural analysis (SA) method to prediagnose the fault and then converts the prediagnosis results into the association graph. Then, the graph and measurements are sent into the GCN model, in which a weight coefficient is introduced to adjust the influence of measurements and the prior knowledge. In this method, the graph structure of GCN is used as a joint point to connect SA based on the model and GCN based on data. In order to verify the effectiveness of the proposed method, an experiment is carried out. The results show that the proposed method, which combines the advantages of both SA and GCN, has better diagnosis results than the existing methods based on common evaluation indicators.

In this paper, a new data-driven fault detection method based on distributed canonical correlation analysis (D-CCA) is proposed to address the plant-wide process monitoring problem. The paper focuses on the distributed plant-wide processes. The core of the proposed method is to reduce uncertainties using correlation information from the neighboring nodes. Furthermore, the cost of data transmission between network nodes is also reduced by the D-CCA algorithm. When the proposed method and the existing methods are compared using the Tennessee Eastman benchmark process, the false alarm rate, fault detection rate, and the detection delay are comparable. This suggests that the proposed method is feasible.

A traction drive control system (TDCS) plays an important role in safety running of high-speed trains. This paper presents a new fault-injection strategy for safety testing and fault diagnosis verification in the TDCS. First, the fault scenarios on the signal level of each faulty component are analyzed. Then, the fault-injection method based on signal conditioning is proposed, and the injected signal, reflecting the fault scenario at a fault point, is generated to simulate the fault scenarios. Subsequently, the injected signal benchmark is constructed for all faults in traction converters, traction motors, sensors, and traction control units. Finally, a fault-injection benchmark platform is developed to simulate various fault scenarios in the TDCS. The simulation and comparison results show that the presented strategy is effective and easy to implement.

This paper proposes a voltage difference residual-based fault diagnosis approach for three-level converters in electric traction systems. Compared to the existing open-circuit fault diagnosis approaches, the proposed approach can, not only detect and identify the faulty devices among the transistors and clamp diodes in both of rectifier and inverter, but also hierarchically diagnose the leg-level and device-level faults. Furthermore, no additional sensors are required and all of specified parameters can be directly identified. Specifically, the current related term in the DC-link voltage difference equations are analyzed and modeled under the normal case and open-circuit fault cases in the transistors and clamp diodes, respectively. Subsequently, the estimation models of the DC-link voltage difference are developed, and then, the residuals and their evaluation functions between estimation and actual are generated. Further, by analyzing the residual evaluation functions for various cases and building the current similarity functions, the hierarchical diagnosis algorithms are proposed to identify the leg-level and device-level faults. Finally, hardware-in-the-loop experiments are performed. The results show the effectiveness and robustness of the proposed approach.