Study On The Technology Of Infrared Absorption Carbon Dioxide Gas Concentration Measurement

The development of science and technology also exert an influence on the natural environment, especially the generation of the greenhouse gases such as carbon dioxide gas. It will not only aggravate the greenhouse effect, but also be harmful to the health of human beings. Therefore, it is very essential to measure the concentration of indoor carbon dioxide gas.This thesis is to introduce the theory of infrared absorption of gas molecular and Lambert-Beer’s law. The hardware circuit and software program of the carbon dioxide gas concentration measurement device has been designed by using non-dispersive electrical modulation measurement structure. In addition, BP neural network is used to calculate the concentrations of carbon dioxide gas. With the adoption of RS485 communication, PC is able to display and store the measurement results which are based on Lab Windows/CVI programming.Carbon dioxide gas concentration measurement device is powered by +5V and it uses the internal A/D module of PIC16F886 microchip, which is with the characteristics of low power consumption, high stability and so on. Microchip outputs pulses to control the modulation of infrared light source. In this way, it can overcome the disadvantages of mechanical modulation, and greatly improve the stability and life of light source. Considering the influence of temperature on the absorption coefficient, this thesis regards the temperature signal with the infrared detector output signals(reference channel signal and measuring signal) as the input layer of BP neural network. Due to the limitation of the processing speed and storage space of microchip, the training process of BP neural network completes in PC by the Matlab software. When getting the structure and parameters of BP network, it can be downloaded to the microchip in order to use the well trained BP network to calculate carbon dioxide gas concentration. Under the premise of ensuring the accuracy of carbon dioxide gas concentration measurement, this method can reduce the cost of the device.The advanced carbon dioxide gas concentration measurement device Model-400 which is made in Germany is regarded as the reference device during the experiment. The results show that using the trained BP neural network with the structure of “three input-single output” to calculate the concentration of carbon dioxide gas, can not only effectively reduce the influence caused by environmental temperature on the measurement results, decrease the measurement error but also improve the accuracy of measurement. The entire device has the advantages of high accuracy, good stability, low power consumption, convenient use and so on. Meanwhile, it provides certain reference for future research on localization and multi parameter gas measurement device of this kind of product.