【作者】 赵楷文； 卢绍文； 赵亮； 于力一； 贾瑶；

【Author】 Kaiwen Zhao;Shaowen Lu;Liang Zhao;Liyi Yu;Yao Jia;State Key Laboratory of Synthetical Automation for Process Industries,Northeastern University;

【机构】 东北大学流程工业综合自动化国家重点实验室；

【摘要】 蒸发过程是拜耳法氧化铝生产工艺中重要的一环,其关键指标是碱液浓度。碱液浓度反应了蒸发过程中蒸水量的多少,间接地反映了氧化铝的生产质量。但碱液浓度无法在线测量。在实际的工业现场中,对碱液浓度的测量一般通过化验室地化验而实现。但化验室地化验需要大量的时间,无法将碱液浓度及时地反馈至控制系统,这样会极大地影响氧化铝的生产质量。为解决上述问题,本文采用基于冶金过程的模拟软件包METSIM设计了氧化铝蒸发过程的模型,并采用基于小波滤波和长短期记忆网络的方法对模型的输出误差进行补偿。利用实际工业现场的数据进行了验证,较为准确地预报了蒸发过程碱液的浓度。表明该模型可以作为运行优化算法和控制算法的实验工具。更多还原

【Abstract】 Evaporation process is an important part of the Bayer alumina production process. Alkali concentration reflects the amount of water evaporating and indirectly reflects the production quality of alumina. But alkali concentration cannot be measured online. In the actual industrial field, the measurement of alkali concentration is generally achieved by laboratory analysis. However, laboratory testing requires a lot of time, so the alkali concentration cannot be fed back to the control system in time, which will greatly affect the production quality of alumina. In order to solve the above problems, the alumina evaporation process model is designed based on chemical process software METSIM and the output error of the model is compensated by the method of wavelet filter and Long-Short Term Network. The concentration of alkali in the evaporation process was predicted accurately by using the data from the industrial site. It is shown that the model can be used as an experimental tool to run optimization algorithm and control algorithm.更多还原