LightGBM-based one-key ore grinding grading control method

By deeply integrating the LightGBM algorithm with the industrial control system, the problems of low control precision and high energy consumption in the grinding system have been solved, realizing precise control and intelligent regulation of the grinding process, and improving production efficiency and stability.

CN120961286AActive Publication Date: 2025-11-18LIAONING ZHONGXIN AUTOMATIC CONTROL
7 Cites 2 Cited by

Patent Information

Application Number
CN202510867172.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-06-26
Publication Date
2025-11-18
Estimated Expiration
2045-06-26

AI Technical Summary

Technical Problem

Existing grinding systems rely on manual experience for adjustment, resulting in low control accuracy, reduced production efficiency, and high energy and material consumption. Traditional PID control is difficult to adapt to multi-variable, strongly coupled, and nonlinear grinding processes. Real-time perception of ore characteristics is insufficient, operation is complex, and the system's intelligence level is inadequate.

Method used

By deeply integrating the LightGBM algorithm with the industrial control system, production data is collected in real time, preprocessed and modeled, a dynamic prediction model is constructed, and deployed to the PLC control layer for millisecond-level real-time decision-making, thereby achieving precise control of the grinding and classification process.

Benefits of technology

It achieves precise control of the grinding process, saves energy and reduces consumption, improves production efficiency and stability, reduces human error, enhances the system's intelligence and adaptability, and reduces energy and material consumption.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to the technical field of dressing plant control, in particular to a LightGBM-based one-key ore grinding grading control method, which comprises the following steps of: preprocessing collected iron ore grinding production data through an industrial internet of things to obtain production data, constructing a data-driven dynamic prediction model by utilizing a LightGBM algorithm, and combining time-space lag cross-correlation and flattening data input to obtain a one-key ore grinding grading control model. And deploying the trained lightweight model to a PLC control layer, and completing optimization and adjustment of corresponding key process parameters. The method has the advantages that by collecting production data in real time and utilizing the LightGBM algorithm to construct the dynamic prediction model, key variables such as the rotating speed and the ore feeding amount of the mill can be accurately and dynamically adjusted, the ore grinding process is optimized, no-load and overload operation of the mill is reduced, and compared with a traditional control method, the ore grinding efficiency is remarkably improved; operators only need to set target parameters, key parameters such as the rotating speed and the ore feeding amount of the mill are optimized in real time through the LightGBM model, it is ensured that the ore grinding fineness stably reaches the standard, and frequent intervention and adjustment are not needed.
Need to check novelty before this filing date? Find Prior Art