矿用红外甲烷传感器温湿度补偿算法研究

Research on Temperature and Humidity Compensation Algorithm for Mine Infrared Methane Sensor

针对环境温湿度对矿用红外甲烷传感器测量精度的影响问题,以人工神经网络为基础,通过遗传优化对传统误差反向传播(back propagation)神经网络算法进行改进,有效改善了BP神经网络收敛速度慢,易陷入局部极小值等缺点。将改进后的算法写入基于MSP430单片机的矿用红外甲烷传感器,在实时甲烷浓度测量中精度提高了4%。

According to the influence of ambient temperature and humidity on the measurement accuracy of mine infrared methane sensor, the back propagation neural network algorithm is improved by genetic optimization based on artificial neural network, which effectively improves the convergence speed of BP neural network slow, easy to fall into the local minimum and other shortcomings. The improved algorithm is written into the mine infrared methane sensor based on MSP430 single chip microcomputer, and the precision is improved by 4% in real-time methane concentration measurement.