Master Thesis MSTR-2011-03

BibliographyNandyala, Rajesh Reddy: Development of a decision support system for dynamic evacuation route planning.
University of Stuttgart, Faculty of Computer Science, Electrical Engineering, and Information Technology, Master Thesis (2011).
87 pages, english.
CR-SchemaF.2.1 (Numerical Algorithms and Problems)
F.2.2 (Nonnumerical Algorithms and Problems)
I.5.5 (Pattern Recognition Implementation)

Evacuation planning is a critical task in disaster emergency management. Numerous types of disasters that happened over the decades in various buildings lead to the development of an array of evacuation planning algorithms. On an event of emergency, traditional evacuation systems notify the occupants about the threat and request the affected occupants to follow the static evacuation paths. The evacuation plans generated by the algorithms lack flexibility in changing the evacuation paths dynamically during evacuation. Occupants are notified to take the static evacuation paths generated, regardless of whatever has happened or is happening in the building. This may lead the occupants into serious problems like leading them to the origin of fire, explosions, collapsed ceilings, blocked exits etc. Moreover, the evacuation plans have no priorities assigned to the occupants. Occupants that are most affected by the disaster are not distinguished and prioritized from the occupants that are not affected by the disaster. To meet these needs, a Priority based Distributed Evacuation Routing (PDER) algorithm is proposed in the thesis. PDER models the building into different areas, the different areas are obtained by the different seperations in the building made by the fire-proof doors installed as per the building regulations. The algorithm generates the evacuation paths parallely for the different areas present in the building, thus reducing the computational time to a great extent. The algorithm assigns priorities to the occupants in different areas, thus giving high priority to the ones that are most affected by the disaster. The current situation in the building is monitored and learned by sensor based monitoring devices and people movement and people count models are derived, which serves as an input to the algorithm. The situation in the building for the next few minutes, the direction of fire and smoke spread is obtained from the Fire Dynamic Simulator (FDS). The algorithm takes into account the simulated information of FDS and the latest available sensor information from the sensor network and generates dynamic evacuation paths, considering the real-time changes in the building. PDER produces sub-optimal evacuation plans with the objective of evacuating the occupants most affected by the disaster at the earliest possible time.

Full text and
other links
PDF (1601778 Bytes)
Access to students' publications restricted to the faculty due to current privacy regulations
Department(s)University of Stuttgart, Institute of Visualisation and Interactive Systems, Visualisation and Interactive Systems
Superviser(s)Weiskopf, Prof. Daniel; Dambier, Michael; Höferlin, Markus
Entry dateJuly 30, 2018
   Publ. Computer Science