基于ELM的智能加药系统的研究

在选矿工艺中,加药过程是影响选矿效果的重要因素,传统浮选加药主要采用人工加药或定量加药方式,其缺点是误差大、效率低和成本高,很难满足选矿厂现代化生产要求。针对这一问题,提出一种基于神经网络和ELM(极限学习机)的控制方案,深入探讨选矿加药的控制方法,并进行严格的试验研究,效果十分理想,对生产实践有重大指导意义。

The Research of Automatic Reagent Feeding System Based on ELM

In the mineral processing industry, the most important factor affecting the quality of mineral production is dosing process. There are two primary ways in traditional floatation dosing, manual or quantitatively dosing. The traditional modes with large error, low efficiency and high cost, can not satisfy the requirements of modern mineral production. In order to solve this problem, a control strategy base on neural network and ELM(Extreme Learning Machine) is proposed. The control strategy of mineral dosing is discussed in detail in the dissertation, and achieveing good results according to the experiment.The research results have important directive to production practices.