BP神经网络在红外CO_2体积分数测量中的应用

Application of BP neural networks in IR CO_2 volume fraction measuring

环境温度对CO2气体体积分数测量的影响是不可忽视的。在CO2体积分数测量中加入温度补偿有助于提高测量装置的精度和有效性,但这很难用传统的数学模型进行温度补偿。反向传播(BP)神经网络特别适用于建立非线性温度补偿网络模型。在实际应用中证明:该方法得到了良好的效果,使CO2气体体积分数测量结果更加准确、稳定。

The effect of environmental temperature on CO2 gas volume fraction measurement cannot be ignored.Add temperature compensation into CO2 gas volume fraction measurement will be helpful to improve precision and effectiveness of measurement device. But it is difficult to compensate temperature using traditional mathematical models. Error back propagation（ BP） neural networks is very suitable to establish nonlinear temperaturecompensation network model. This method is proved to have a good effect in application,and make the results of CO2 volume fraction measurement more accurate and stable.