Research On Calibration Algorithm For Air Quality Monitoring Data

In recent years,China has made rapid changes in air pollution control.and air grid monitoring has played an active role in pollution control.However,the large number of miniaturized monitoring equipment used in grid monitoring has a number of limitations including large measurement error,many equipment abnormal conditions,time drift,and environmental mutations.At the same time,the potential value contained in the historical data of large-scale environmental monitoring has not been effectively tapped.The limitations caused by calibrating miniaturized monitoring equipment through historical monitoring data are critical scientific issues to be solved urgently.Concerning this scientific problem,this paper develops an air quality monitoring data calibration algorithm based on evidential reasoning theory and fuzzy cognitive map.The main work of this thesis is summarized as follows:(1)Aiming at the problem that the correlation of air pollutants is difficult to be captured and the samples data are sparse,we propose an air quality monitoring data prediction algorithm based on fuzzy cognitive map.First,we use the historical monitoring data of the mobile sensor to generate an initial fuzzy cognitive map.Secondly,we develop a scheme to dynamically optimize the FCM using the global search characteristics of the real code genetic algorithm,through the error function.Finally,according to the constructed air quality data prediction model combined with the reasoning mechanism of the fuzzy cognitive map,the prediction of air quality is realized.Experimental results show that the method can accurately predict the air quality monitoring data and has a lower time complexity than other algorithms,and at the same time provides technical support for air pollution prevention and control.(2)Aiming at the problem that monitoring data are easy to drift and their calibration baseline are not reasonable,we propose an integrated calibration algorithm based on evidential reasoning theory and fuzzy cognitive map.First,we use the evidential reasoning theory to fuse the optimal fuzzy cognitive maps generated by multiple mobile sensors to get a higher-level fuzzy cognitive map integration model.Secondly,we combine the inference mechanism of the integrated model and fuzzy cognitive map to generate the monitoring benchmark of low-cost sensors.Finally,the abnormal data is detected by quantifying the difference between the single model and the integrated model,and then the reference data is used to replace the abnormal data to achieve the self-calibration of mobile sensor data.The model can achieve segmented calibration of abnormal air quality data without the supervision of professional sites,providing a reliable data guarantee for the treatment of air pollution.