火灾自动探测的复合特定趋势算法

Multiple Input Special Trend Algorithm for Automatic Fire Detection

本文提出了二输入信号的复合特定趋势检测算法。它能够根据信号的稳定值识别复合信号位于稳定值上方的正向变化和稳定值下方的负向变化趋势。在趋势判断函数中引入判别门限，有效地克服了随机噪声干扰；在火灾信号确认中使用了“趋势持续”方法。通过标准试验火和不同环境下非火灾信号的模拟测试表明，该算法探测火灾信号及时，可靠且具有极低的误报警率。

In this paper,a two-input detective algorithm for the special trend is proposed.There are the steady state values in fire senrors.The detector can distinguish the positive trend above the steady state value and the nagative trend below it.A threshold is introduced to the calculation of the special trend so that noise be overcome. The trend duration algorithm is used for the determination of fire signals.The detector has been tested by different fire and no fire signals. The results show that the algorithm discovers the test fires in time and has a very low false alarm rate.