基于单模式对应矫正的近红外甲烷探测仪设计

Near Infrared Methane Detector Design Based on Single Mode Corresponding Correction

甲烷气体对红外辐射有选择性吸收的关系,利用这一性质可以设计红外强度映射检测甲烷浓度,目前的甲烷探测仪中红外吸收谱线的谐波幅度与浓度会受到环境中噪声的影响发生线性漂移,影响了甲烷探测的准确度与系统的稳定性;提出基于神经网络线性矫正的近红外甲烷探测仪,使用PIC16F877单片机实现甲烷探测的控制核心处理器,完成信号的采集与浓度计算声光报警等功能,设计信号调理电路对采集的信号进行外围处理,在单片机中实现信号谐波幅度与浓度的线性矫正;实际的探测仪测试中,通过曲线拟合与误差分析证明这种探测仪的甲烷探测准确率高达97%,具有很强的实用性。

Methane gas to infrared radiation selective absorption of the relationship, the use of this nature can design infrared intensity map detection methane concentration. The methane detector of infrared absorption spectrum line harmonic amplitude and concentration will be affected by the influence of the noise in the environment happened linear drift, the impact of the methane detection accuracy and stability of the system. Based on neural network linear correction of near infrared methane detector, the use of PIC16F877 single--chip microcomput er to realize the control of methane detection core processor, complete signal acquisition and concentration calculation acousto--optic alarm functions, design and signal conditioning circuit to collect signals, the microcontroller peripheral processing realize signal harmonic amplitude and concentration of the linear correction. The actual detector test, through the curve fitting and error analysis shows that this kind of detector of methane detection accuracy reaches as high as 97%, have very strong practicability.