30 September 2024 in IEEE Signal Processing Letters
Pedestrian intrusion detection is crucial to ensuring safe railway operation. Current pedestrian detection algorithms lack consideration for real-world railway scenarios, such as the reflective properties of screen doors and train windows, may mistakenly trigger pedestrian intrusion alerts. Scale variability and pedestrian overlap often lead to detection inaccuracy, making them inadequate for addressing the specific requirements of railway perimeter security. This letter introduces an innovative pedestrian detection algorithm that incorporates Mirror Translation Attention (MTA) and Feature Pooling Enhancement (FPE). MTA, including mirror flipping and offsetting the feature mapping, could significantly mitigate missed detection caused by reflective surfaces. Additionally, we introduce sparsity to the inputs of the self-attention, which significantly enhancing the model's inference speed. A multi-scale approach is adopted to accommodate the diversity in pedestrian sizes, while the FPE addresses occlusion issues across various scales. Compared to the advanced YOLOv8 model, the proposed method improves AP50 by 1.6% to 92.11% and reduces model parameters by 63.55% in our self-built railway pedestrian intrusion dataset.
6 February 2026 in IEEE Access
Load modeling is difficult due to the uncertainty and time-varying properties of the load. Through the recently proposed ambient signals load modeling approach, these properties can be more frequently tracked. However, the large dataset of load modeling results becomes a new problem. In this paper, a hierarchical temporal and spatial clustering method for load models is proposed, after which the large load model dataset can be represented by several representative load models (RLMs). In the temporal clustering stage, the RLMs of one load bus are chosen through clustering to represent all the load models of the load bus at different times. In the spatial clustering stage, the RLMs of all the load buses form a new set, and the RLMs of the system are chosen through clustering. In this way, the large sets of load models are represented by a small number of RLMs, through which the storage space for the load models is significantly reduced. The validation results in the IEEE 39 bus system have shown that the simulation accuracy can still be maintained after replacing the load models with the RLMs. In this way, the effectiveness of the proposed hierarchical clustering framework is validated.