Developing accurate kinetic models to predict cracking processes is essential for enhancing production quality and efficiency. Establishing a comprehensive free radical network that precisely predicts the product distribution in the cracking processes of all light hydrocarbons presents significant challenges. This study introduces a reaction network initiated from four individual hydrocarbons using an automatic reaction network generator, RMG. We construct a detailed network by merging the networks of individual hydrocarbons to accommodate the unique reaction pathways of different feeds. The model's accuracy is significantly improved by identifying key reactions through sensitivity analysis and refining their kinetic parameters with the PSO algorithm. This optimized kinetic model is implemented within a one-dimensional plug flow reactor (PFR) developed in Matlab to simulate the cracking process under various feeds and operating conditions. The model accurately predicts yield distributions of key products along the reactor length, demonstrating good agreement with experimental data, even for minor products.

中文翻译：

---

基于自动反应网络生成的唯一和混合轻烃蒸汽裂解的最优动力学建模


开发精确的动力学模型来预测开裂过程对于提高生产质量和效率至关重要。建立一个全面的自由基网络，精确预测所有轻烃裂解过程中的产物分布，这带来了重大挑战。本研究介绍了一个使用自动反应网络发生器 RMG 由四个单独的碳氢化合物引发的反应网络。我们通过合并单个碳氢化合物的网络来构建一个详细的网络，以适应不同原料的独特反应途径。通过灵敏度分析识别关键反应并使用 PSO 算法优化其动力学参数，该模型的准确性得到了显著提高。这种优化的动力学模型是在 Matlab 开发的一维活塞流反应器 （PFR） 中实现的，用于模拟各种进料和操作条件下的裂解过程。该模型准确预测了沿反应器长度的关键产品的产量分布，证明与实验数据具有良好的一致性，即使是次要产品也是如此。