Predicting cobalt ion concentration in hydrometallurgy zinc process using data decomposition and machine learning.

Area of Science:

• Environmental Science
• Chemical Engineering
• Data Science

Background:

• Solid waste significantly contributes to environmental pollution.
• Effective management of solid waste, particularly in industrial processes like zinc hydrometallurgy, is crucial.
• Cobalt ion concentration in zinc purification effluent directly impacts water pollution and resource efficiency.

Purpose of the Study:

• To develop a hybrid prediction model for accurate cobalt ion concentration monitoring in zinc hydrometallurgy effluent.
• To address the limitations of manual testing, including time, cost, and delays.
• To mitigate water pollution and optimize zinc powder usage in industrial purification.

Main Methods:

• A hybrid model combining data decomposition and machine learning algorithms was proposed.
• Ablation and contrast experiments were conducted using consistent training and test datasets.
• Quantitative metrics analysis, including root mean square error and coefficient of determination, was performed.

Main Results:

• The hybrid prediction model demonstrated the smallest error and best fit for cobalt ion concentration.
• Significant reductions in error metrics were observed: RMSE (4.2%-73.9%), MAE (7.1%-93.4%), MPE (7.7%-86.7%).
• The coefficient of determination improved by 1.3%-134.6%, indicating enhanced predictive accuracy.

Conclusions:

• The hybrid model effectively predicts cobalt ion concentration, preventing water pollution from industrial zinc purification.
• The model offers practical significance for technicians to control zinc powder input, reducing resource waste and production costs.
• This approach enhances environmental management in industrial hydrometallurgy processes.