【摘要】 提出了一种新型的、应用多传感器数据融合技术于当前高大空间建筑的火灾探测报警系统的方法。该种方法采用多传感器探测火灾多种参量,先使用神经-模糊网络进行局部判定,局部判定结果与融合中心进行历史数据的融合,得到误差后确定下一步全局决策。通过试验与仿真证明此方法大大提高了火灾探测的及时性,同时又提高系统可靠性,减少了误报警。更多还原

【Abstract】 A novel design of fire detection system based on multi-sensors data fusion technology for high and large-span space,is introduced. This method apply neuro-fuzzy network on diverse types of fire detecting signals to make local decisions, then modify the fusion center’s decision rule with the error between the result of the local decision and the data of the fusion center’s decision . Results of experiments and simulations show that this method can improve advanced-warning and veracity effectively, and reduce mis-warning significantly.更多还原