Research on optimization of control parameters of gravity shaking table

Area of Science:

• Mineral Processing
• Machine Vision
• Data-Driven Modeling

Background:

• Shaking table performance relies on understanding internal control parameters and external ore zone characteristics.
• Accurate analysis of processing indicators and control parameter mapping is crucial for efficient ore beneficiation.

Purpose of the Study:

• To develop an adaptive optimization method for shaking table control parameters to maximize beneficiation efficiency.
• To construct a data-driven model characterizing the relationship between shaking table parameters and ore properties.

Main Methods:

• Utilizing image processing and machine vision to extract features from ore belt images.
• Employing a visual experimental system to gather multi-scale zone characteristic information.
• Developing and optimizing support vector regression (SVR) models, including sparrow search algorithm optimized SVR (SSA-SVR).

Main Results:

• Experimental data on ore belt geometric characteristics met statistical distribution requirements.
• The SSA-SVR model demonstrated superior performance, overcoming data limitations and meeting industrial needs.
• The proposed method achieved continuous optimization of multiple shaking table control parameters.

Conclusions:

• The developed adaptive optimization method effectively enhances shaking table beneficiation efficiency.
• The SSA-SVR model provides a robust solution for optimizing complex separation processes.
• This approach offers a reliable method for continuous, data-driven optimization in mineral processing.