基于GA-BP神经网络的红外CO_(2)传感器湿度补偿研究

Research on Humidity Compensation of Infrared CO_(2) Sensor Based on GA-BP Neural Network

为了提高红外CO_(2)气体传感器的探测灵敏度和精度,首先研究了不同镀膜对非色散扁锥腔CO_(2)气体传感器的红外吸收效率和灵敏度的影响.然后搭建了湿度实验平台,着重研究了环境湿度对气体浓度测量结果的影响.最后,采用遗传算法优化的BP神经网络算法(GA-BP)对传感器进行了湿度补偿.实验结果表明:在室温条件下、0~2000×10^(-6)浓度范围内,镀金腔体的CO_(2)传感器具有更高的红外吸收效率和灵敏度;在40%~80%湿度范围内,CO_(2)气体传感器的测量误差与相对湿度密切相关,最高误差达645×10^(-6).采用GA-BP算法数据融合补偿后,传感器湿度漂移得到了较好抑制,整体平均误差小于±110×10^(-6),表明CO_(2)气体传感器的测量精度得到了大幅提升.

In order to improve the detection sensitivity and accuracy of the infrared CO_(2) gas sensor,the effect of different coatings on the infrared absorption efficiency and sensitivity of the nondispersive flat cone CO_(2) gas sensor was first studied.Then a humidity experiment platform was built,and the influence of environmental humidity on gas measurement results was emphatically studied.Finally,the genetic algorithm optimized BP neural network algorithm(GA-BP)was applied to the humidity compensation of the sensor.The experimental results show that at room temperature,the CO_(2) sensor with goldplated cavity has higher infrared absorption efficiency and sensitivity in the concentration range of 0~2000×10^(-6).The measurement error of the gas sensor is closely related to the relative humidity within the range of 40%to 80%,and its maximum error can reach 645×10^(-6).After GA-BP algorithm data fusion compensation,the humidity drift of the sensor is better suppressed.The overall average error is less than±110×10^(-6),indicating that the measurement accuracy of the CO_(2) gas sensor has been greatly improved.