Traditional indoor navigation algorithms generally only consider the geometrical information of indoor space. However, the environmental information and semantic parameters of a fire are also important for evacuation routing in the case of a fire. It is difficult for traditional indoor navigation algorithms to dynamically find an indoor path when a fire develops. To address this problem, we developed a multi-semantic constrained three-dimensional (3D) indoor fire evacuation routing method that considers multi-dimensional indoor fire scene-related semantics, such as path accessibility, path recognition degree, and fire parameters. Our method enhances the navigation semantics of indoor space by extending the fire-related components of indoor model based on IndoorGML and integrating location semantics of IndoorLocationGML. We also propose quantifiable indoor fire-oriented routing semantics and establish a navigation cost function that evaluates semantic changes during a fire. We designed an indoor routing algorithm with multiple semantic constraints based on the A* algorithm. The indoor routing results were analyzed and compared in simulation experiments. The experimental results showed that the proposed model can remove unusable nodes and edges from the obtained navigation path and provides a safer and more effective evacuation route than traditional algorithms.

中文翻译：

---

三维室内消防疏散通道

传统的室内导航算法通常只考虑室内空间的几何信息。但是，火灾的环境信息和语义参数对于火灾情况下的疏散路线也很重要。对于传统的室内导航算法，当发生火灾时，很难动态地找到室内路径。为了解决这个问题，我们开发了一种多义约束三维（3D）室内火灾疏散路由方法，该方法考虑了多维室内火灾场景相关的语义，例如路径可访问性，路径识别度和火灾参数。我们的方法通过扩展基于IndoorGML的室内模型中与火有关的组件并集成IndoorLocationGML的位置语义来增强室内空间的导航语义。我们还提出了可量化的面向室内火灾的路由语义，并建立了评估火灾期间语义变化的导航成本函数。我们基于A *算法设计了具有多个语义约束的室内路由算法。室内路由结果在仿真实验中进行了分析和比较。实验结果表明，与传统算法相比，该模型可以从获得的导航路径中去除不可用的节点和边缘，并提供了更安全，更有效的疏散路线。室内路由结果在仿真实验中进行了分析和比较。实验结果表明，与传统算法相比，该模型可以从获得的导航路径中去除不可用的节点和边缘，并提供了更安全，更有效的疏散路线。室内路由结果在仿真实验中进行了分析和比较。实验结果表明，与传统算法相比，该模型可以从获得的导航路径中去除不可用的节点和边缘，并提供了更安全，更有效的疏散路线。