Abstract Theoretically predicted rate coefficient parameters play an increasing role in combustion kinetic models. Understanding the uncertainty propagation during RRKM/master equation modeling and quantitatively evaluating the uncertainties of the phenomenological rate coefficients would be a great boon to chemical modeling. In the present work, the global uncertainty and sensitivity analysis were performed for a prototypical multiwell and multichannel reaction system (on the C4H7 potential energy surface) over the temperature and pressure range of 300–2300 K and 0.01–100 atm. Six reactions including chemical activation reactions and unimolecular reactions were studied, focusing on the uncertainty of both absolute rate coefficients and their branching ratios. To reveal the nature of uncertainty propagation from input parameters at different theoretical levels, five scenarios were designed to mimic the uncertainty from high-, moderate- and low- levels of theories. For absolute rate coefficients, the temperature dependence of uncertainties for all the reactions is found to be much stronger than the pressure dependence. The pressure dependence is mainly determined by the reaction type (chemical activation or thermal dissociation, directly-connected or well-skipping). For the branching ratio, sensitivity analysis shows that the input parameters that are very important in determining the absolute rate coefficients do not necessarily have high sensitivity coefficients in predicting the branching ratio. As a consequence, the uncertainty of the theoretically calculated branching ratio can be smaller than that of the rate coefficients.

中文翻译：

---

典型多井多通道反应系统RRKM/主方程建模的全局不确定性分析

摘要 理论预测的速率系数参数在燃烧动力学模型中扮演着越来越重要的角色。了解 RRKM/主方程建模过程中的不确定性传播并定量评估现象学速率系数的不确定性将是化学建模的一大福音。在目前的工作中，在 300–2300 K 和 0.01–100 atm 的温度和压力范围内，对原型多孔和多通道反应系统（在 C4H7 势能表面上）进行了全局不确定性和灵敏度分析。研究了包括化学活化反应和单分子反应在内的六个反应，重点研究了绝对速率系数及其支化率的不确定性。为了揭示来自不同理论水平的输入参数的不确定性传播的性质，设计了五种情景来模拟来自高、中和低水平理论的不确定性。对于绝对速率系数，发现所有反应的不确定性的温度依赖性比压力依赖性强得多。压力依赖性主要由反应类型（化学活化或热解离、直接连接或井跳过）决定。对于支化率，敏感性分析表明，在确定绝对率系数方面非常重要的输入参数在预测支化率时不一定具有高灵敏度系数。作为结果，