Abstract
Fire evacuation simulation is used to simulate the fire evacuation procedures by involving human-like agents. In this paper, the fire evacuation simulation is designed and developed by adopting the BDI agent plug-in. BDI (Belief, Desires, Intentions) is a technique used in modelling the multi-agent system. A tool and BDI methodology are introduced to help in modelling human behaviour and the decision making of an agent. In this paper, the usability of the BDI methodology and BDI agent plug-in tool is studied through a case study of a fire evacuation environment. The case study covers the three main components needed in a fire evacuation simulation: the fire (the spread of the fire and smoke), the building layout (the classroom and physical objects), and the human-like multi-agents. Using the Unity game engine, a fire evacuation simulation system is built based on the requirements, methodology, and design. The usability of the BDI agent plug-in tool can be proven by observing the results of the fire evacuation simulation and the reaction of agents when encountering the fire situation. However, there are also some limitations of this fire evacuation simulation. Therefore, there are works to be done to develop a more realistic fire evacuation simulation and more human-like multi-agents in future.
Multi-agents system for breast tumour detection in mammography by deep learning pre-processing and watershed segmentation
