| 摘要:基于有机胺吸收剂的化学吸收法在船舶碳捕集与封存技术中最具潜力，但对吸收剂降解状态的监控较为困难。通过真实试验数据拟合，发现乙醇胺的真实降解过程与“一般混合效应退化模型”相吻合。进一步通过“间接监督学习”模型将易检测的塔顶多种排气排放物浓度融合起来，构建可实时反映吸收剂降解状态的健康指数。研究发现健康指数比各原始传感器数据更能反映吸收剂的真实降解状态，其中吸收塔健康指数对训练样本吸收剂状态诊断平均相对误差为2.38%，对真实试验吸收剂状态诊断平均相对误差为5.14%，可有效反映有机胺吸收剂的真实降解状态。 |
| 关键词: 船舶CCS系统 有机胺吸收剂 状态诊断 退化模型 数据融合 气体传感器 |
|
| State Diagnosis of Absorbent for Ship Flue Gas Decarbonization Based on Multi-Sensor Fusion Method |
| HUANG Hao,SHEN Feixiang,LI Xiaobo,LI Ke,ZHANG Yan |
| Shanghai Marine Diesel Engine Research Institute, Shanghai 201108, China;Shanghai Marine Diesel Engine Research Institute, Shanghai 201108, China;National Engineering Research Center of Special Equipment and Power System for Ship and Marine Engineering,Shanghai 200090,China |
| Abstract:The chemical absorption method based on organic amine absorbents is the most promising technology in marine carbon capture and storage（CCS）technologies，but it is difficult to monitor the degradation state of absorbents.Through the fitting of test data，it was found that the degradation process of MEA was consistent with the "general mixed effect degradation model".The “indirect supervised learning” model was used to realize data fusion of the concentrations of various easily-detectable exhaust emissions at the top of tower to construct a health index（HI）that could reflect the degradation state of the absorbent in real time.The study found that：HI could better reflect the true degradation state of the absorbent than the original individual sensor data.Especially，the average relative error of the HI-absorption in the diagnosis of the absorbent state of the training sample was 2.38%，and the average relative error of the diagnosis of absorbent state in actual test was 5.14%，so HI could effectively reflect the true degradation state of organic amine absorbents. |
| Key words: marine CCS system organic amine absorbent status diagnosis degradation model data fusion gas sensor |
