权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Online Estimation of Catalyst Activity in a Packed bed Reactor

填充床反应器中催化剂活性的在线估算

基本信息

批准号：
15560661
负责人：
MATSUMOTO Shigeru
金额：
$ 1.98万
依托单位：
Tohoku University
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (C)
财政年份：
2003
资助国家：
日本
起止时间：
2003 至 2004
项目状态：
已结题

项目摘要

A state observer which was not influenced by observation noise was designed to estimate the concentration and the temperature profiles in a packed-bed reactor. By use of orthogonal collocation method, nonlinear ordinary differential equations (ODE) were derived from the non-dimensional mass and the heat balance model of the reactor. Extended Kalman filter based on the derived ODE model was used to estimate both the state variables and frequency factor of rate constant.Simulation was performed to confirm the estimation performances of the designed observer. The results showed that the observer estimated the concentration and the temperature profiles accurately even in the presence of the observation noise, and the frequency factor also agreed with the value of the simulator.Next, experiment of the perfect ethylene oxidation was carried out by using the packed-bed reactor (90cm length, 2.5cm diameter). The concentration and the temperature profiles and the frequency factor were estimated well by using measured temperature of 3 to 5 points in the reactor and concentration measured by infrared gas analyzer. The estimated frequency factor agreed with the value obtained by the preliminary experiment within the accuracy of 10%. Since the catalyst deactivation with the progress in the reaction time was not observed, catalyst deactivation rate was not detected by using the observer. However, because this observer can estimate the frequency factor online, it is considered that the catalyst deactivation can be also estimated by watching the estimated value of the frequency factor with the progress in the reaction time.

设计了一个不受观测噪声影响的状态观测器来估计填料床反应器内的浓度和温度分布。利用正交配点法，对反应器的无因次质量和热平衡模型进行了非线性常微分方程推导。在导出ODE模型的基础上，利用扩展卡尔曼滤波对状态变量和速率常数的频率因子进行估计。通过仿真验证了所设计观测器的估计性能。结果表明，在存在观测噪声的情况下，观测器仍能准确地估计出浓度和温度曲线，且频率因子与模拟器的值基本一致。其次，采用填料床反应器（长90cm，直径2.5cm）进行乙烯完美氧化实验。利用反应器内3 ~ 5点测得的温度和红外气体分析仪测得的浓度，可以很好地估计出浓度、温度分布和频率因子。估计的频率因子与初步实验值吻合，精度在10%以内。由于没有观察到催化剂的失活随反应时间的进展，因此没有使用观察者来检测催化剂的失活率。但是，由于该观察者可以在线估计频率因子，因此可以认为，通过观察频率因子随反应时间的进展的估计值，也可以估计催化剂的失活。