Accurate short-term forecasts of passenger flow in metro systems under delay conditions are crucial for emergency response and service recovery, which pose significant challenges and are currently under-researched. Due to the rare occurrence of delay events, the limited sample size under delay condictions make it difficult for conventional models to effectively capture the complex impacts of delays on passenger flow, resulting in low forecasting accuracy. Recognizing the strengths of large language models (LLMs) in few-shot learning due to their powerful pre-training, contextual understanding, ability to perform zero-shot and few-shot reasoning, to address the issues that effectively generalize and adapt with minimal data, we propose a passenger flow forecasting framework under delay conditions that synthesizes an LLM with carefully designed prompt engineering. By Refining prompt design, we enable the LLM to understand delay event information and the pattern from historical passenger flow data, thus overcoming the challenges of passenger flow forecasting under delay conditions. The propmpt engineering in the framework consists of two main stages: systematic prompt generation and prompt refinement. In the prompt generation stage, multi-source data is transformed into descriptive texts understandable by the LLM and stored. In the prompt refinement stage, we employ the multidimensional Chain of Thought (CoT) method to refine the prompts. We verify the proposed framework by conducting experiments using real-world datasets specifically targeting passenger flow forecasting under delay conditions of Shenzhen metro in China. The experimental results demonstrate that the proposed model performs particularly well in forecasting passenger flow under delay conditions.
Train delays can propagate rapidly throughout the Urban Rail Transit (URT) network under networked operation conditions, posing significant challenges to operational departments. Accurately predicting passenger travel choices under train delays can provide interpretable insights into the redistribution of passenger flow, offering crucial decision support for emergency response and service recovery. However, the diversity of travel choices due to passenger heterogeneity and the sparsity of delay events leads to issues of data sparsity and sample imbalance in the travel choices dataset under metro delays. It is challenging to model this problem using traditional machine learning approaches, which typically rely on large, balanced datasets. Given the strengths of large language models (LLMs) in text processing, understanding, and their capabilities in small-sample and even zero-shot learning, this paper proposes a novel Passenger Travel Choice prediction framework under metro delays with the Large Language Model (DelayPTC-LLM). The well-designed prompting engineering is developed to guide the LLM in making and rationalizing predictions about travel choices, taking into account passenger heterogeneity and features of the delay events. Utilizing real-world data from Shenzhen Metro, including Automated Fare Collection (AFC) data and detailed delay logs, a comparative analysis of DelayPTC-LLM with traditional prediction models demonstrates the superior capability of LLMs in handling complex, sparse datasets commonly encountered under disruption of transportation systems. The results validate the advantages of DelayPTC-LLM in terms of predictive accuracy and its potential to provide actionable insights for big traffic data.
Delays can adversely affect passengers, operational efficiency, and punctuality, necessitating effective methods to identify and analyse stations vulnerable to such delays. This study proposes a three-stage analytical framework. In the first stage, the 3σ rule is utilized to define normal passenger volume ranges and establish a specific time window, focusing on periods significantly affected by delays. Next, a multivariate time series clustering method is proposed to identify stations with stable demand and high volume, considering passenger volume differences both among and within stations. In the final stage, the effects of delays on these key stations are assessed by examining duration, starting, and ending times, and passenger volume variation, providing a comprehensive analysis of delay impact. The proposed method is validated on a real-world case: the 2021 delay incident at Shanghai Metro's Longyang Road Station. Our case study revealed that stations significantly impacted by delays are not limited to the specific line or direction of the delay, but also include opposite direction and network transfer stations. The starting time of effects at stations occurs in three stages—prior to, during, and after the delay—with full recovery at most key stations not occurring until the third time window. Additionally, both increases and decreases in passenger volumes due to the delay present considerable implications. Our results aid transit managers in better managing delays, thereby improving passenger satisfaction and operational efficiency. This paper also offers a comprehensive framework and an empirical study, providing valuable guidance for future research in this area.
Transit delays can adversely affect passengers, operational efficiency, and daily lives. It is important to develop effective methods to identify and analyze train stations vulnerable to delays. This paper proposes a three-stage analytical framework for analyzing train station delays. In the first stage, the 3-sigma rule defines normal passenger volume ranges and establishes a time window affected by delays. Next, a multivariate time series clustering method identifies stations with stable demand and high volume, considering passenger volume differences both among and within stations. In the final stage, the effects of delays on these key stations are assessed by examining starting, duration, and ending times, and passenger volume variation, providing a comprehensive analysis of delay impact. The proposed framework is illustrated using two real-world incidents: the 2021 delay incident at Longyang Road Station of Shanghai Metro and the 2019 delay incident on the Taoyuan–Luohu section of Shenzhen Metro. Case studies revealed that affected stations are not limited to the specific line or direction of the delay, but also include opposite-direction and transfer stations. Station impacts exhibit phased onset and recovery patterns. Additionally, both increases and decreases in passenger volumes due to the delay present considerable implications. While both incidents exhibit common propagation and recovery patterns, the Shanghai incident displays wider passenger impacts and longer recovery periods, whereas the Shenzhen incident exhibits narrower impacts and faster recovery. Our results will aid transit managers in better managing delays, thereby improving passenger satisfaction and operational efficiency. This paper also offers an integrated station-level analytical framework and initial cross-case empirical evidence, while broader validation remains needed.
Equipment fault prediction in maglev train systems poses substantial challenges: fault events are inherently rare, class distributions are severely imbalanced, and individual equipment units are subject to complex spatial and functional couplings that single-device statistical approaches fundamentally cannot capture. To address these challenges, this study proposes a Knowledge Graph-driven Graph Neural Network (KG-GNN) framework. A fault knowledge graph encompassing equipment, fault, temporal, and environmental entities is constructed to unify multi-source maintenance data. Graph connectivity is established via three spatial relation types (co-location, co-zone, and co-level), with edge weights derived from Laplacian-smoothed Lift scores quantifying fault co-occurrence strength. A two-layer GATv2Conv-based graph attention network is designed: the first layer employs four-head attention with explicit edge-weight integration to capture heterogeneous neighborhood influences, while the second layer produces compact node embeddings via single-head attention. A Top-20 sparsification strategy suppresses weak-association noise, and training under severe class imbalance is stabilized through Focal Loss and F2-Score-guided early stopping. On the test set, the proposed method achieves an F2-Score of 0.5703, Recall of 0.6825, and AUC-ROC of 0.9329 (single-run evaluation); multi-seed evaluation (5 seeds) yields F2 = 0.5645 ± 0.0035, Recall = 0.6789 ± 0.0095, and AUC-ROC = 0.9298 ± 0.0026, outperforming the MLP baseline by 18.3% in F2-Score and substantially exceeding GCN (F2 = 0.1476 ± 0.0176) and GATConv (F2 = 0.4284 ± 0.0097). Ablation studies confirm the individual contributions of authentic graph topology, precise edge weighting, and graph sparsification to overall performance.