In  order to quickly  and accurately  predict the  evacuation time of passenger ships perso nnel, an e vacu ation simu-latio n mo del with good calculati on acc urac y wa s combined with a BP artifi ci al neura l ne twork m odel with h igh calculation effic ien cy, and a metho d for quickly pr edictin g th e ev acuation time of pas se nger ships was p roposed. B as ed on the simula-tion r esul ts of pa ssenger ships evacu ation i n d iffer ent scenari-os, a BP neural netwo rk predict ion model for the eva cuation time of p asse nger ships person nel was c onstructed. The results sho w th at the prop osed me thod has high cal culation eff iciency, an d t he prediction re lative e rror of differ ent evacuation sc enar-ios is 0. 9% ~ 3. 6%, which can provide decision-making basis and new technical tools for passenger ship emergency manage-ment.