Research On Error Compensation Method For Remote Sensing Measurement Of Mobile Pollution Source

Mobile pollution sources refer to non-stationary,mobile air pollution sources,including motor vehicles,mobile construction machinery,ships and aircraft,and other vehicles that emit a large number of harmful gases while driving.Remote sensing is a method for measuring vehicle emissions in traveling by remote sensing based on optical principle.Compared with traditional methods,such as idle speed method and operating condition method,remote sensing method can quickly measure the pollutant concentration of vehicles passing through the measuring site,and has the advantages of high testing efficiency,high degree of automation and low test cost.Small impact on traffic and other advantages.Because the remote sensing method adopts the open detection method,the measurement process is easily affected by the complex environment around the road(such as high temperature and high humidity,strong wind and so on).At the same time,the exhaust gas will diffuse to form plume,and the surrounding environmental factors,especially the local air flow,will have a great effect on the exhaust plume.Aiming at the problem that the remote sensing detection technology of pollutant gas from mobile pollution source is vulnerable to external environmental interference,this paper combines the transfer entropy correlation causality analysis and adaptive fusion estimation method.A new error compensation method based on transfer entropy and adaptive Kalman filter is proposed for remote sensing of mobile source emissions.Using the method,the measurement error prediction model under multiple interferences is first established by extreme learning machine,the actual measurement process is then transformed into the multi-sensor virtual observation model whichl is used to achieve the multi-sequence decomposition of the original observation sequence,and finally the multiple virtual observation sequences are reconstructed by adaptive Kalman filter fusion.In the process of reconstruction,transfer entropy is used to represent the multi-disturbance unbalance measurement and optimize the observation noise covariance coefficient in adaptive Kalman filter.Experimental results illustrate the effectiveness of the proposed method.