Related Experiment Video
Updated: Aug 28, 2025
06:52
Lab-Scale Model to Evaluate Odor and Gas Concentrations Emitted by Deep Bedded Pack Manure
Published on: July 19, 2018
6.4K
Simulation, prediction and optimization of typical heavy metals immobilization in swine manure composting by using
Hao-Nan Guo1, Hong-Tao Liu2, Shubiao Wu3
1Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, 100101, China; College of Resources and Environment, University of Chinese Academy of Sciences, Beijing, 100049, China.
Journal of Environmental Management
|September 22, 2022
Summary
Machine learning models effectively predicted and optimized heavy metal immobilization during composting. Gradient boosting regression identified key factors, achieving high immobilization rates for Cu, Zn, Cd, As, and Cr.
Area of Science:
- Environmental Science
- Agricultural Engineering
- Computational Science
Background:
- Traditional composting experiments face limitations in analyzing complex data.
- Heavy metal contamination in livestock manure compost poses environmental risks.
- Machine learning (ML) offers advanced data analysis capabilities for environmental remediation.
Purpose of the Study:
- To predict and optimize heavy metal immobilization during composting using ML models.
- To identify key factors influencing heavy metal bioavailability and immobilization.
- To assess the efficacy of ML in managing heavy metal risks in compost.
Main Methods:
- Integration of four ML models: multi-layer perceptron regression, support vector regression, decision tree regression, and gradient boosting regression.
- Utilized genetic algorithm for optimizing heavy metal immobilization predictions.
- Performed feature importance analysis using gradient boosting regression.
Main Results:
- Gradient boosting regression demonstrated superior performance in predicting heavy metal bioavailability and immobilization.
- Initial heavy metal bioavailability, total phosphorus, and composting duration were identified as primary factors influencing bioavailability variations (>75%).
- Optimized immobilization rates reached up to 79.53% for Cu, 31.30% for Zn, 14.91% for Cd, 46.25% for As, and 66.27% for Cr.
Conclusions:
- ML, particularly gradient boosting regression, is a powerful tool for predicting and optimizing heavy metal immobilization in composting.
- The study highlights the potential of ML for effective risk management of heavy metals in livestock manure composting.
- Findings suggest ML can enhance the safety and environmental sustainability of composted materials.
