Abstract: In order to improve the scientific nature of crew's emergency decision when dealing with cabin emergencies, cabin emergency decision model and emergency decision process based on CBR (Case-based Reasoning) theory is proposed in this paper. Attribute weight is obtained by rough set theory, and an algorithm which combine structure similarity, attribute similarity and global similarity is adapt to receive the global optimal similarity case and its emergency response plan. The cabin emergency response system is designed based on this model, and a target case is analyzed by using this model, and an experimental validation of the emergency response to lithium battery fires in passenger cabin luggage racks and seats is conducted. The study shows that, after the use of the model, the efficiency of emergency response increases by 22.28% and 18.96% respectively, which can verify the scientific and efficacy of the model.

Key words: cabin emergency;emergency decision-making, case-based reasoning, rough set theory, case representation