Proceedings of the International Conference on Industrial Design and Environmental Engineering, IDEE 2023, November 24–26, 2023, Zhengzhou, China

Research Article

Identification Method of Urban Atmospheric Particulate Pollution Sources Based on Energy Spectrum Characteristics and Neural Network

Download60 downloads
  • @INPROCEEDINGS{10.4108/eai.24-11-2023.2343482,
        author={Mingming  Wang and Xiang  Zhang and Zhou  Zhou},
        title={Identification Method of Urban Atmospheric  Particulate Pollution Sources Based on Energy  Spectrum Characteristics and Neural Network},
        proceedings={Proceedings of the International Conference on Industrial Design and Environmental Engineering, IDEE 2023, November 24--26, 2023, Zhengzhou, China},
        publisher={EAI},
        proceedings_a={IDEE},
        year={2024},
        month={2},
        keywords={energy spectrum characteristics; neural network technology; atmospheric  particulate matter; pollution source; pollution location; identification method},
        doi={10.4108/eai.24-11-2023.2343482}
    }
    
  • Mingming Wang
    Xiang Zhang
    Zhou Zhou
    Year: 2024
    Identification Method of Urban Atmospheric Particulate Pollution Sources Based on Energy Spectrum Characteristics and Neural Network
    IDEE
    EAI
    DOI: 10.4108/eai.24-11-2023.2343482
Mingming Wang1, Xiang Zhang1,*, Zhou Zhou2
  • 1: Nantong Ecological Environment Monitoring Center
  • 2: Jiangsu Suli Environmental Technology Co., Ltd.
*Contact email: ntssthjj12369@163.com

Abstract

Most of the pollution source identification mechanisms are set on the target structure, resulting in low identification efficiency and long response time for unit identification. To this end, a design and analysis method for identifying urban atmospheric particulate pollution sources based on energy spectrum features and neural networks has been proposed. According to the current experimental requirements, basic identification indicators for atmospheric particulate matter pollution sources are set, multi-level methods are used to improve recognition efficiency, a multi-level cross pollution source recognition mechanism is constructed, and an energy spectrum feature+neural network atmospheric particulate matter pollution source recognition model is designed. Urban pollution source recognition is achieved through distributed locking processing. The experimental results show that the five selected areas for identifying atmospheric particulate matter pollution sources have a response time of less than 0.3 seconds for final unit identification, which has good practical application effects. With the assistance and support of energy spectrum characteristics and neural networks, it has high pertinence and practical application value.