中文摘要：

银催化剂是乙烯选择氧化成环氧乙烷的唯一有效催化剂，描述银催化剂进行乙烯氧化反应机理和动力学研究一直是环氧乙烷合成领域的热点。本文针对环氧乙烷合成反应的银催化剂失活问题，采用支持向量机对催化剂失活进行建模，同时，根据催化剂失活的特性，考虑了温度、时间对催化剂影响的先验知识，在实验室模型的基础上针对工业装置建立了环氧乙烷合成反应器的催化剂失活模型，并通过模型对环氧乙烷合成反应进行了计算。计算结果表明，基于支持向量机的环氧乙烷合成反应催化剂失活模型的预测误差在2％以内。与传统的建模方法相比，基于支持向量机的催化剂失活模型能够满足生产中对环氧乙烷反应器优化时对模型的要求。

英文摘要：

Sliver is the only valid catalyst of ethylene oxide synthesis. The description of the function of silver catalyst in reaction mechanism is a hotspot in this domain. In order to solve the nonlinear problem of catalyst deactivation of ethylene oxide synthesis, support vector regression was proposed. Taking account of prior knowledge of influence of temperature and time, a model was developed based on industrial data. The industrial production process was calculated and it shows that the prediction error of the model is less than 2%. Compared to conventional modeling method, catalyst deactivation model based on SVR can be used for optimization of ethylene oxide reactor.