Knowledge Representation Model for Evolution of Crowd Evacuation Stability

In large-scale crowd evacuation, it is necessary to make reasonable and quick decision to organize the crowd evacuation based on the large amount of crowd evacuation data. Obviously, knowledge representation for the evolutionary mechanism of crowd evacuation stability is the first step during the crowd evacuation decision-making. In this paper, we propose a novel knowledge representation model for the evolution mechanism of crowd stability state. We use the principle of self-organizing mapping network to discretize the characteristics of crowd evacuation scenarios and stability state evolution. Using discretized variables, a two-dimensional condition-decision information table is formed based on the representation paradigm of rough set. Furthermore, the hidden relationship between the features of crowd trampling scenarios and the evolutionary characteristics of the stability state can be more effectively analyzed and obtained, which provides scientific basis for the intelligent decision-making of crowd evacuation.