Search

Anomaly detection is essential to ensure metro vehicles' safe operation. Error reconstruction-based anomaly detection methods have been widely studied because they only need to be trained by normal data and do not require much anomaly data, which is challenging to obtain. However, sometimes the auto-encoder network for error reconstructing “generalizes” so well that it also rebuilds the anomaly well, leading to missed anomaly detection. Therefore, this paper proposes an undercarriage image-driven anomaly detection method for metro vehicles based on adversarial memory enhancement. Firstly, this study performs component segmentation based on YOLOv5 detection results and constructs a component anomaly detection dataset. Secondly, an anomaly detection method based on memory enhancement and adversarial training of encoding-decoding-encoding structure is proposed for component anomaly detection. It enables the auto-encoder to reconstruct the image better. Thirdly, the combined indicator of the difference between potential features and reconstruction error is used as an anomaly indicator for anomaly detection of metro components, reducing the rate of fault misses. The experimental results on the established dataset demonstrate that the proposed method reduces false negative rates of 92.4%, 92.6%, 74.6%, and 59.1% compared with AEL2, AESSIM, GANomaly, and MemAE, respectively.

With the increase in running speed, the aerodynamic characteristics of high-speed trains have a significant impact on running stability, energy consumption and passenger comfort. Since the shape of the high-speed train head can directly influence the surrounding airflow, optimizing the head shape is the primary way to improve the aerodynamic performance of the train. This paper reviews current research studies on the surrogate-based aerodynamic shape optimization of high-speed train heads, aiming to provide a comprehensive reference for designers to enhance design efficiency and optimization performance. The entire optimization process is divided into four essential steps, and the key optimization technologies in each step are discussed, including parametric modeling, computational fluid dynamics (CFD) simulation, surrogate model and optimization algorithm. By introducing the practical applications of these technologies, we summarize their advantages and disadvantages and suggest four potential research directions for the future.