Pedestrian Flow Modeling in Obstacle Environment Based on Inverse Reinforcement Learning
Bibliographic record
Abstract
Developing accurate pedestrian flow simulation models is crucial for enhancing the safety and operational efficiency of high-traffic environments such as subway stations, railway terminals, and airports. Traditional pedestrian flow modeling approaches typically rely on manually defined behavioral rules, which suffer from strong subjectivity, lack of standardization, and difficulty in parameter calibration. With advancements in data acquisition and computational technologies, data-driven and machine learning-based modeling methods have gradually emerged. Among them, inverse reinforcement learning (IRL) has garnered attention due to its ability to learn behavioral strategies directly from real trajectories without the need for predefined reward functions. This study proposes a pedestrian flow modeling method based on IRL, focusing on environments with fixed obstacles. Real-world video observations were collected, and high-precision pedestrian trajectory data were extracted to construct a nested simulation model integrating both forward and inverse reinforcement learning. The model defines generalized pedestrian state and action spaces and uses joint training of the policy and reward networks to precisely capture and simulate pedestrian behavior. Simulation results demonstrate that the proposed model closely aligns with real- world observations. Compared to the traditional Social Force Model, it exhibits superior trajectory fitting accuracy and speed distribution consistency, indicating strong adaptability and simulation credibility. This provides a novel approach to modeling pedestrian behavior in complex environments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".