Design Of Real-Time Online Monitoring And Processing System For Fixed Source Carbon Emission Data

The emissions of CO₂ greenhouse gases pose significant threats to the natural environment and human health.The industrial sectors,such as thermal power generation and steelmaking,which have fixed pollution source exhaust CO₂emissions,are the main sources of greenhouse gas production.Therefore,developing a precise and real-time online monitoring system with fixed pollution source exhaust CO₂ emission data is particularly urgent.This paper designs a high-precision non-dispersive infrared CO₂ gas sensor with a new optical structure,pre-processing of gas path environmental conditions,and temperature and humidity automatic compensation function,which can perform on-site real-time detection of CO₂ monitoring system.The gas sensor in the system is designed with an optical structure that has a Fresnel focusing light source and a carbon nanotube black body coating.ZEMAX optical software was used for sequential mode simulation to determine the Fresnel concentrator parameters and sensor structure size.In non-sequential mode,the system’s new optical structure was simulated with three reference-type optical structures,including conventional aluminum-plated optics,aluminum-plated optics with Fresnel focusing light source,and carbon nanotube black body coating optics,and the results showed that the system structure can effectively enhance the total light irradiance of the detector surface and improve radiation uniformity.In FLUENT fluid dynamics software,energy radiation simulation was carried out to study the relationship between sensor radiation intensity and inlet flow velocity on the detector response efficiency.It was found that the higher the radiation intensity,the higher the detector radiation response efficiency;when the inlet flow velocity was controlled at around 10 m/s,the detector response efficiency was the highest.Based on the above conclusions,the system hardware and software modules were designed.The circuit structure and driver program of each peripheral device in the system were designed based on Cortex-A7.The human-machine interaction interface was designed on the Qt Create platform,and the system data was managed using the SQLite database to construct the entire system’s working hardware and software conditions.Based on the above hardware and software foundation,calibration tests were carried out on the system and reference structure sensors on the experimental platform,and the results showed that the system structure sensors had the best fitting effect and sensitivity.Then,static performance tests were performed on the system sensors,and they all achieved the expected results.Finally,environmental temperature and humidity interference experiments were conducted on the sensors,and it was found that the sensor had serious temperature and humidity drift phenomenon,and the highest relative error was as high as 25.6%.Complete temperature and humidity drift testing on the experimental platform to provide data support for the testing and verification of compensation algorithms.The results show that the relative error value of the sensor compensated by the radial basis function（RBF）neural network algorithm is reduced to 3.5%;The RBF neural network algorithm optimized using particle swarm optimization（PSO）maintains the relative error of the sensor below 1.9%after compensation,which has better compensation effect.