基于分布式感知的铝电解智能识别系统及应用  

Intelligent identification system and application of aluminum electrolysis based on distributed perception

作　　者：邹忠 彭晨 李劼 张红亮 罗若木 肖胜华 韩博 ZOU Zhong;PENG Chen;LI Jie;ZHANG Hongliang;LUO Ruomu;XIAO Shenghua;HAN Bo(School of Metallurgy and Environment,Central South University,Changsha 410083,China;Hunan ALWIT Science and Technology Co.Ltd.,Changsha 410083,China)

机构地区：[1]中南大学冶金与环境学院,湖南长沙410083 [2]湖南阿尔惠特科技股份有限公司,湖南长沙410083

出　　处：《中南大学学报（自然科学版）》2023年第2期732-743,共12页Journal of Central South University:Science and Technology

基　　金：国家重点研发计划项目(2022YFB3304902);云南省重大科技专项计划项目(202202AB080017);国家自然科学基金联合项目(U2202253);国家自然科学基金资助项目(62133016,51974373);中南大学前沿交叉项目(2023QYJC007);湖南省研究生科研创新项目(CX20220223)。

摘　　要：传统铝电解过程人工干预较多,决策的随意性和不可靠性直接影响着电解槽的运行稳定性,进而对技术经济指标产生重大影响。为推进铝电解生产实现智能制造转型升级,采用新型分布式检测装置实现电解槽状态参数实时获取,并在此基础上结合两级智能识别方法,实现生产过程的精细化控制。研究结果表明:在实际投运过程中,上述技术使国内某铝电解厂350 kA系列阳极效应系数降低到0.006次/(槽·d),平均电流效率提高1.3%,每生产1 tAl节电338 kW·h,取得了显著的增效节能减排效果,可为铝电解槽数字孪生体的构建提供参考。Traditional aluminum electrolysis leans primarily on humans, and meanwhile the cell’s stability and economic performance are also directly influenced by the unreliability of man-made decisions. In order to encourage intelligent manufacturing transformation, a new type of distributed detection device and gathered realtime data on electrolytic cell status parameters were implemented, a two-level intelligent identification system was combined with this data to realize effective production control. The results show that with the use of these technologies in operation, the anode effect coefficient falls to 0.006 times per cell a day, the current efficiency increases by 1.3% on average, and the energy consumption reduces by 338 kW·h per 1 t aluminum of 350 kA series, which offers a significant energy-efficient effect and also lays the foundation for the construction of the digital twin of the aluminum electrolytic cell.

关 键 词：铝电解 分布式检测 在线检测 智能识别 

分 类 号：TF821[冶金工程—有色金属冶金]