基于粒子群神经网络算法的交通枢纽火灾应急疏散评价

在构建交通枢纽火灾应急疏散评价系统的基础上，建立用粒子群优化算法（PS0）和BP神经网络相结合的混合算法来进行综合评价的模型。该模型全面考虑影响应急疏散的各种因素，将定性分析和定量分析相结合，使评价结果更符合实际、更可靠。实例计算结果表明：PSO-BP方法简便、通用，其结果符合运营实际，能很好的反应交通枢纽火灾状况下应急疏散的实施效果。

Evaluation of Fire Emergency Evacuation of Traffic Hub Based on PSO-BP Model

On the basis of evaluating fire emergency evacuation of traffic hub, the authors present the model of the mixed algorithm of PSO（particle swarm optimization） algorithm and BP（back-propagation）network for comprehensive evaluation, which could fully consider factors of emergency evacuation. Exemplified calculation result shows that PSO-BP model is simple, convenient and universal. The result accords with the reality of operating, the implementing effect of fire emergency evacuation of traffic hub that it can be very good..