Adaptive compensation for measurement error in remote sensing of mobile source emissions

Abstract Using remote sensing technology to monitor mobile pollution sources is an advanced technology to prevent air pollution. This paper proposes an error compensation model that can prevent remote sensing from being subjected to complex and variable environmental disturbances. Using this novel method, the measurement error prediction model under multiple interferences is established by Extreme Learning Machine. Then, the actual measurement process is transformed into the multi-sensor virtual observation model that is used to achieve the sequence decomposition of the original sequence. Finally, the fusion of virtual observation sequences is performed by the Adaptive Kalman Filter. Transfer Entropy is used to represent the multi-disturbance unbalance measurement and optimize the observation noise covariance coefficient in Adaptive Kalman Filter. Experimental results showed that compared with traditional method, our model performed better. The results indicated that our method can quickly and effectively compensate the measurement error and improve the environmental adaptability of the instrument.