敞口体系燃爆状态多传感器联合辨识方法研究

Research on Deflagration Identification Method of Multi-Sensor in Open System

针对敞口体系燃爆辨识人工观测存在安全风险、单一传感器容易误判问题,提出一种基于光强和温度的多传感器联合燃爆辨识方法。采用光敏传感器阵列增大有效检测面积,避免单一传感器检测的偶然性;使用自适应加权算法处理光强冗余数据,消除背景光强变化带来的干扰;将光强和温度的特征数据提取后进行初步决策和构造概率分配函数,经由D-S证据理论进行决策层融合选择最可信的初步决策作为辨识结果。在嵌入式平台上对辨识方法进行实现,实验表明多传感器联合燃爆辨识方法的辨识准确率高达99.4%,实现了燃爆自动化检测。

The manual observation for open system deflagration identification has a high risk in safety,and it’s more likely to misjudge while using single sensor. To solve the problem,a method of multi-sensor joint deflagration identification based on light intensity and temperature sensing is introduced. A photosensitive sensor array was used to increase the effective detection area and avoid the accidental detection by a single sensor;the adaptive weighted algorithm was used to process the redundant data of light intensity to eliminate the interference caused by the background light intensity change;the characteristic data of light intensity and temperature were extracted to make preliminary decision and construct the probability distribution function. After the decision level fusion using the DS evidence theory,the most reliable preliminary decision was selected as the identification result. The identification method was verified on the embedded system. The experiment showed that the identification accuracy rate of the multisensor deflagration identification method was as high as 99.4%,and the automatic identification of deflagration was achieved.