塑料作为人类现代日常生活和工业应用中不可或缺的材料，其生产量在过去50年中呈指数级增长，导致废弃塑料也呈现爆炸式的增长。以聚烯烃为主，难以生物降解的废弃塑料的积累已成为全球性问题，其资源化回收利用迫在眉睫。论文利用热解实验研究聚烯烃塑料的转化规律和反应机理，设计两段法热解工艺实现高效回收烯烃单体；同时，以结构导向集总法（SOL）为基础对聚烯烃热解过程构建分子管理模型，实现热解过程模拟并指导工艺设计和过程优化。论文设计了非等温和等温快速热解实验，分析聚烯烃热解产物分布并推断反应机理。聚乙烯（PE）和聚丙烯（PP）低温热解主要生成碳链分布广泛的饱和脂肪烃和不饱和脂肪烃，反应机理存在随机断裂和链端断裂的竞争。基于反应机理，提出了随机断裂机理动力学模型，成功模拟了高密度聚乙烯（HDPE）的等温热解过程，证明了大分子聚合物链长的变短过程受随机断裂调控。进一步，从机理出发，将相变过程和气相反应分离，成功设计两段法热解制低碳烯烃。第一段低温热解（500 °C），生成以蜡和油为主的产物，直接作为第二段高温热解（700-1000 °C）的原料，进而生成低碳烯烃。PE和PP单独热解实验结果，在二段热解800 °C条件下，低碳烯烃产率最高分别为76.1 wt.%和71.8 wt.%。研究表明，低温和高温下热解分别由随机断裂和链端断裂机制主导，存在分子间和分子内氢转移的竞争。进一步，实验还拓展至混合PE/PP以及废弃口罩的热解处理。从分子水平动力学方法出发，基于聚烯烃分子量的泊松分布模型、SOL方法和符合机理的反应规则设计，首先对PP热解过程建立了自动生成反应网络的分子管理模型。由于建立了区分一次反应和二次反应的动力学网络，模型确定的速率常数符合阿伦尼乌斯方程，预测结果与实验数据具有很好的一致性；该方法也同样适用于PE热解过程。针对两段热解工艺，构建了低温-高温热解分子管理模型，实现了对HDPE在不同温度组合条件下两段热解行为的模拟，可分析特定产物的生成和分解历程。模型预测不仅量化了温度对热解的重要作用以及两种机理的竞争关系，还显示出模型的指导作用和优化热解工艺的潜力。最后，对聚氯乙烯（PVC）热解脱氯过程也初步构建了分子管理模型，实现了低温条件下PVC脱氯模拟。

As indispensable materials for human daily life and industrial applications, the production of plastics has increased exponentially over the last 50 years, resulting in an explosive growth of waste plastics. The accumulation of non-biodegradable waste plastics, mainly waste polyolefins, has become a concern, and their recycling is imminent. In this work, the degradation and reaction mechanisms of polyolefins are studied by pyrolysis experiments, and a two-stage pyrolysis process is designed to achieve a high yield of light olefins. Then, a molecular management model is established for the pyrolysis of polyolefins based on the structure-oriented lumping method (SOL) to realize the simulation of the pyrolysis and guide the optimization of the process.This work designs non-isothermal and isothermal pyrolysis experiments to analyze the product distribution and deduce the reaction mechanism. The pyrolysis of polyethylene (PE) and polypropylene (PP) mainly produces saturated and unsaturated aliphatic hydrocarbons with a wide carbon number distribution at low temperature, and there is competition between random scission and chain-end scission. Based on the analysis of the reaction mechanism, a kinetic model of the random scission mechanism is proposed, and the isothermal pyrolysis process of HDPE is successfully simulated, demonstrating that the degradation of polymers is regulated by random scission.Further, a two-stage pyrolysis process is designed for producing light olefins based on the reaction mechanism and the separation of phase change and gas-phase reactions. The evolution of light olefins (C2–C4) during the pyrolysis is investigated at a constant condition in the first stage (500 °C) and various conditions in the second stage (700–1000 °C). In the second stage, the highest light olefins yield of PE and PP is 76.1 wt.% and 71.8 wt.% at 800 °C, respectively. It is also proposed that the random scission and chain-end scission mechanisms dominate at low and temperatures, respectively, and there exists competition between intermolecular and intramolecular hydrogen transfer. In addition, the two-stage pyrolysis is also successfully applied to the mixed PE/PP and waste masks.Based on the Poisson distribution model, the SOL method, and the design of reaction rules consistent with the reaction mechanism, a molecular management model with automatic reaction network generation is established for polypropylene pyrolysis. Due to the establishment of a kinetic reaction network that distinguishes between primary and secondary reactions, the rate constants determined by the model are consistent with the Arrhenius equation, and the predictions are in good agreement with the experimental data. In addition, the model is also applicable to PE pyrolysis. The molecular management model for the two-stage pyrolysis enables the simulation of HDPE pyrolysis at different temperatures and tracks the generation of specific products. The model predictions quantify the important role of temperature and the competition between two mechanisms, revealing the guidance of the model to optimize the pyrolysis process. Finally, a molecular management model for polyvinyl chloride (PVC) dichlorination is constructed and realizes the simulation of PVC dichlorination at low temperature.