Dioxin emission modeling using feature selection and simplified DFR with residual error fitting for the grate-based...

Area of Science:

• Environmental Science
• Chemical Engineering
• Data Science

Background:

• Municipal solid waste incineration (MSWI) is a key waste-to-energy technology in China.
• Dioxin (DXN) emissions are critical environmental indicators for MSWI process optimization.
• Developing high-precision, fast DXN emission models for operational control is challenging.

Purpose of the Study:

• To develop a novel, accurate, and fast method for measuring dioxin emissions in MSWI.
• To enable real-time operational optimization and control of MSWI processes.
• To address the difficulty in constructing precise DXN emission models.

Main Methods:

• Utilized simplified deep forest regression with residual error fitting (SDFR-ref).
• Optimally reduced high-dimensional process variables using mutual information and significance tests.
• Employed a gradient enhancement strategy with residual error fitting for improved performance.

Main Results:

• The SDFR-ref method demonstrated superior measurement accuracy compared to other approaches.
• The proposed method significantly reduced time consumption in DXN emission measurement.
• Validation using a real-world dataset from Beijing's MSWI plant confirmed effectiveness.

Conclusions:

• The SDFR-ref method provides a highly accurate and efficient solution for DXN emission monitoring.
• This approach facilitates better operational optimization and environmental control in MSWI.
• The study highlights the potential of advanced machine learning for environmental applications.