Atmospheric particulate matter pollution has attracted much wider attention globally. In recent years, the development of atmospheric particle collection techniques has put forwards new demands on the real-time source apportionments techniques. Such demands are summarized, in this paper, as how to set up new restraints in apportionment and how to develop a non-linear regression model to process complicated circumstances, such as the existence of secondary source and similar source. In this study, we firstly analyze the possible and potential restraints in single particle source apportionment, then propose a novel three-step self-feedback long short-term memory (SF-LSTM) network for approximating the source contribution. The proposed deep learning neural network includes three modules, as generation, scoring and refining, and regeneration modules. Benefited from the scoring modules, SF-LSTM implants four loss functions representing four restraints to be followed in the apportionment, meanwhile, the regeneration module calculates the source contribution in a non-linear way. The results show that the model outperforms the conventional regression methods in the overall performance of the four evaluation indicators (residual sum of squares, stability, sparsity, negativity) for the restraints. Additionally, in short time-resolution analyzing, SF-LSTM provides better results under the restraint of stability.
Keywords: Particle source apportionment; Regression; Self-feedback LSTM network; Time series.
Copyright © 2021. Published by Elsevier B.V.