In order to effectively depict the pedestrian evacuation process, based on the cellular automaton, a novel evacuation model was proposed by electromagnetism mechanism and fuzzy neural network. Combined the static factor and dynamic factor, the calculation formula of transition probability and evacuation strategy were given. Meanwhile, the optimization function is given with evacuation time, average system velocity and flow rate, and it is solved by electromagnetism mechanism and fuzzy neural network. At last, experiments were conducted with the simulation platform to study the relationships of evacuation time, exit width as well as initial pedestrian density. The results showed that evacuation time had a negative correlation with exit width, and an appropriate increasing for average system velocity would reduce the evacuation time.