2019 年 68 巻 3 号 p. 278-284
Various preparations for possible disasters should be made so as to minimize the damage in the disasters such as a fire or an earthquake. Several kinds of disaster simulations have been developed to recognize the impact of the damages. In particular, a lot of evacuation simulation methods using multi-agent system, which are considered as the useful and powerful method to deal with complex system containing a lot of elements, have been proposed for guiding evacuees safely. The Social force model is one of the evacuation simulation models that consider the influences and interaction of the other evacuees. In order to improve the precision of the simulation, the parameters included in it should be appropriately determined. In this paper the determination method of the parameters in the social force model is proposed, in which the evolutional computing method is applied for the collecting evacuees flow data.