论文部分内容阅读
针对硫浮选过程中常规检测方法难以准确检测浮选槽液位的缺陷,提出一种基于相关向量机(RVM)的浮选液位软测量方法。该方法基于采集的浮选泡沫表层图像,通过提取硫浮选泡沫溢流速度和泡沫稳定度动态图像特征,融合浮选过程充气量、矿浆流量等过程参数,结合RVM建模思想,实现硫浮选过程中浮选槽液位的预测。工业数据仿真结果验证了所提方法的有效性、可行性。
Aiming at the defect that it is difficult to detect the liquid level of the flotation tank by the conventional detection method in the process of sulfur flotation, a soft sensor based on correlation vector machine (RVM) is proposed. Based on the collected flotation froth surface images, through the extraction of dynamic image characteristics of sulfur foam flotation foam velocity and foam stability, flotation process influx, flow rate and other process parameters, combined with the RVM modeling ideas to achieve sulfur floating Selection of flotation tank level during the forecast. Industrial data simulation results verify the effectiveness and feasibility of the proposed method.