A cognitive-based routing algorithm for crowd dynamics under incomplete or even incorrect map knowledge.

The article focuses on the development and validation of a cognitive-based routing algorithm for crowd simulations, which allows pedestrians to navigate environments with varying levels of map knowledge, including incomplete or outdated information. Using a hospital floor plan as a test scenario, the study introduces concepts such as local form for personalized network perception, a fold line based distance computation method to better represent room geometry, and a span-tree based algorithm to determine next steps when map knowledge is partial. Numerical experiments demonstrate that reasonable evacuation routes can be found even with limited knowledge, though pedestrians with zero or minimal map awareness tend to traverse more rooms, potentially increasing evacuation time. The research highlights the trade-off between computational complexity and realistic individual and collective routing behaviors, while acknowledging limitations such as the assumption of ideal memory and the need for further studies on memory imperfections, environmental aids, and diverse real-world topologies.