摘要
针对传统选煤厂煤泥压滤过程中受人为操作影响较大的问题,开展了压滤系统控制模型的构建及压滤过程流量阈值的优化。基于自适应BP神经网络模型搭建了数据处理平台,形成了从料浆物性到工作参数的神经网络结构模型,探究了模型对滤饼含水量预测的准确性。利用微小流量检测技术实现了压滤机压滤结束状态自判断,结合压滤效果及压滤效率确定了阈值流量,并进一步分析了压滤过程中滤水流量的变化规律。结果表明:逻辑判断优化后,实际滤饼水分与目标水分的绝对误差为±1.5%,当阈值流量为0.05m3/h时,压滤后煤泥平均含水量为24.0%。改造后,滤饼含水量的极差由4.5%减小至1.3%,运行作业更加稳定。此外,压滤时间均在15min以下,提高了压滤过程的工作效率,实现了选煤厂的降本增效。
As the pressure filtration process of coal slime in traditional coal preparation plant is greatly affected by human operation, the construction of the control model of the filter press system and the optimization of the flow threshold in the filter press process were carried out. Based on the adaptive BP neural network model, the data processing platform was built, and the neural network structure model from slurry physical properties to working parameters was formed, and the accuracy of the model in predicting the water content of filter cake was explored. The micro-flow detection technology was used to realize the self-judgment of the end state of the press. The threshold flow was determined by the filter performance, and the variation of the filter flow rate in the process of filter pressure was further analyzed. The results showed that the absolute error between the actual cake moisture and the target moisture was ±1.5% after the optimization of logic judgment. When the threshold flow was 0.05m~3/h, the average water content of the slime after the filter press was 24%. After transformation, the range of water content of filter cake was reduced from 4.5% to 1.3%, and the operation was more stable. In addition, the filter pressing time was less than 15min, which improved the working efficiency of the filter pressing process and realized the cost reduction and efficiency increase of the coal preparation plant.
作者
杨鹏民
YANG Peng-min(China Coal Shaanxi Yulin Energy and Chemical Co.,Ltd.,Yulin 719000,China)
出处
《煤炭工程》
北大核心
2023年第1期18-22,共5页
Coal Engineering
关键词
煤泥压滤
微小流量检测
选煤厂智能化
coal slime filter press
micro flow detection
intelligentization of coal preparation plant
