基于预测模型的浮选过程pH值控制

pH control in flotation process based on prediction model

矿浆pH值是泡沫浮选过程中的一个非常重要的被控量.目前,多数选厂的矿浆pH值控制基本是依靠现场工人定期对矿浆样本进行pH值测量,凭主观经验对pH调整剂进行调整.由于操作工人的主观性和随意性的影响以及矿浆样本pH值测量与药剂调整间存在的较长的时间滞后,矿浆pH值波动频繁,很难使矿物浮选保持在一个稳定最优生产状态下运行.为了使矿浆pH值保持在一个期望的生产状态,基于浮选泡沫表面视觉信息提出了一种新的矿浆pH值控制方法,分别采用基于泡沫视觉信息的自适应遗传混合神经网络AG–HNN和自适应遗传PID(AG–PID)控制方法建立了矿浆pH值预测模型和pH值控制模型,基于所建立预测和控制模型对浮选药剂用量进行调整,解决了浮选矿浆pH值波动问题.工业浮选现场的实验结果表明该方法可以使矿浆pH值保持在一个期望的范围内,有效提高浮选性能.

The pulp pH value is an important controlled variable for process operation in mineral flotation process,which is controlled through regularly adjusting reagent dosage by experienced workers according to the pH measurement results of the current pulp samples.However,due to the influence of the subjectivity and casualness of workers and the long time-lag between the pH measurement of the pulp sample and the reagent adjustment,the pH value of the pulp fluctuates frequently,causing unstable and undesirable production states for mineral flotation.A novel pH control method is proposed to maintain the pulp pH value at an expected production state based on the froth visual information.Firstly,a pH prediction model and a pH control model are built by froth visual information based on the adaptive genetic hybrid neural network（AG–HNN） and adaptive genetic PID（AG–PID） control method,respectively;and then the dosage is adjusted according to the pH value prediction model and the pH control model.Several industrial-scale experiments demonstrate that this method can maintain the pH value of the pulp in an expected range and consequently improve the flotation performance effectively.