To improve the efficiency and accuracy of building fire detection and prediction,the decentralized mode of edge computing was used to enhance the intelligent analysis capability of building terminals,and the edge decentralized fire inver-sion and situation prediction method were proposed.The results show that for the dynamic fire scenarios,both the decentral-ized inversion and prediction methods can produce relatively accurate results at the order of 10-1 s and achieve the discrimi-nation for all 34 groups of common false alarm scenarios.Meanwhile,the decentralized mode improves the scalability of inver-sion and prediction methods for different building structures.The research results can provide decision support for the emergency rescue and personnel evacuation of building fires.
关键词
建筑火灾/参数反演/态势预测/模拟仿真
Key words
building fire/parameter inversion/situation prediction/simulation
维普
万方数据