Title :
Predicting evacuation time using learnable building block method
Author :
Teekul, Jaksuthep ; Sinthupinyo, Sukree
Author_Institution :
Dept. of Comput. Eng., Chulalongkorn Univ., Bangkok, Thailand
Abstract :
Evacuation simulator is an important tool which can create solution for emergency circumstance. Evacuation simulation becomes more realistic but it has to tradeoff between memory consumption and processing time. This paper proposes a method to predict evacuation time by converting floor layout to building block structure and predicts evacuation time of each building block. Then we can use the predicted result to calculate total evacuation time of scenario without re-simulate whole scenario. The proposed method can reduce the processing time and memory usage which make user get total evacuation time faster but still maintain correctness of result.
Keywords :
building management systems; emergency services; learning (artificial intelligence); multilayer perceptrons; software agents; agent-based simulation; emergency circumstance; evacuation simulator; evacuation time prediction; floor layout; learnable building block method; memory consumption; memory usage reduction; processing time reduction; Data models; Fires; Floors; Layout; Predictive models; Solid modeling; agent-based simulation; building; building blocks; evacuation time; muiti-layers perceptron; prediction;
Conference_Titel :
Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON), 2012 9th International Conference on
Conference_Location :
Phetchaburi
Print_ISBN :
978-1-4673-2026-9
DOI :
10.1109/ECTICon.2012.6254277