重介分选过程产品指标在线预测方法研究

On-line prediction of product indicators in dense medium coal separation

重介质选煤过程中,灰分值的实时在线检测能够有效提高控制系统的性能,但是当下灰分值难以在线检测,针对这一问题,首先对重介质选煤工艺过程进行了简要概括,并分析了其主要的工业过程变量,进一步构建基于LightGBM(Light Gradient Boosting Machine)的重介质选煤过程灰分预测模型。通过实际的工业过程数据对模型预测结果进行验证,实验结果表明,此模型能够较准确地预测灰分大小,对重介质选煤过程灰分预测的研究起到参考作用。

In the process of heavy medium coal preparation,the real-time online detection of ash content can prominently enhance the performance of the control system,while the online detection of ash content is difficult.Aiming at this problem,firstly,the process of heavy medium coal separation is briefly summarized,and the main industrial process variables are analyzed,according to which the ash content prediction model of heavy medium coal separation process based on LightGBM(Light Gradient Boosting Machine)is built and tested.The test results show that the model can predict the ash content accurately,which can offer reference for the ash content prediction of heavy medium coal preparation.