Abstract: In order to improve the effectiveness of crowd behavior simulation, this paper improves the social force model based on the characteristics of crowd behavior. Firstly, in view of the problem that the speed jump will occur during the simulation of social force model is inconsistent with the actual situation, the social force model is modified considering the actual situation of human movement and combining with the human bearing capacity. Secondly, aiming at the phenomenon of multiple collisions and velocity oscillations in simulation, fuzzy rules are used to optimize the personnel range coefficient in social force model, so as to give full play to the subjective initiative of personnel. The results show that the optimized model can effectively reduce the number of collisions and velocity oscillation. Finally, based on the analysis of the above optimization results, a hybrid- modified model is established based on restrictive conditions and fuzzy rules.

Key words: fire control, social force model, fuzzy rule, relative speed, restrictive conditions