分子水平动力学模型和机器学习方法相结合研究废弃塑料热解

doi:10.11949/0438-1157.20240585

摘要/Abstract

摘要：

随着全球塑料产量的增加，废弃塑料处理问题日益严重，热解技术作为一种将废弃塑料转化为高附加值产品的方法，引起了广泛关注。通过分子水平动力学模型与机器学习相结合的方法，研究了聚乙烯（PE）的热解过程。首先，利用分子水平动力学模型针对不同分子量分布的PE原料，生成大规模热解数据集。然后，基于大规模数据集构建了9种机器学习模型，评估其预测能力及特征重要性，分析影响热解产物收率的关键因素。结果表明，反应时间和热解温度是主要影响因素，KNN模型在气、液相产物预测中表现最佳。通过优化机器学习模型和扩大数据集，可以显著提升热解过程的预测准确性和效率，为废弃塑料资源化提供了新的思路和方法。

关键词: 热解, 机器学习, 分子水平动力学模型, 废弃塑料, 模拟, 废物处理

Abstract:

With the increase of global plastics production, the problem of waste plastics disposal has become increasingly serious. Pyrolysis technology has attracted widespread attention as a method to convert waste plastics into high value-added products. The pyrolysis process of polyethylene (PE) was investigated by combining molecular-level kinetic model with machine learning methods. First, a large-scale pyrolysis dataset was generated for PE raw materials with different molecular weight distributions using molecular level kinetic model. Then, 9 machine learning models were constructed based on the large-scale dataset to evaluate their predictive ability and feature importance, and analyze the key factors affecting the product yields. The results show that reaction time and pyrolysis temperature are the main factors, and the KNN model performs the best in the prediction of gas and liquid phase products. The study also demonstrates that the simulation accuracy and efficiency of the pyrolysis process can be significantly improved by optimizing the machine learning model and expanding the dataset, which provides new ideas and methods for waste plastics resourcing.

Key words: pyrolysis, machine learning, molecular level kinetic model, waste plastics, simulation, waste treatment

中图分类号:

TQ 028.8

王茂先, 孙启典, 付哲, 华放, 纪晔, 程易. 分子水平动力学模型和机器学习方法相结合研究废弃塑料热解[J]. 化工学报, 2024, 75(11): 4320-4332.

Maoxian WANG, Qidian SUN, Zhe FU, Fang HUA, Ye JI, Yi CHENG. Understanding pyrolysis process of polyethylene by combined method of molecular-level kinetic model with machine learning[J]. CIESC Journal, 2024, 75(11): 4320-4332.

图/表 13

图1 分子水平动力学模型与机器学习结合的研究方法示意图

Fig.1 Schematic diagram of combined method of molecular level kinetic model with machine learning

图2 多种PE原料分子量分布

Fig.2 Schematic of molecular weight distribution of various PE raw materials

表1 PE原料分子量分布参数

Table 1 Molecular weight distribution parameters of PE raw materials

编号	a	b	c	d	f	M	nums	numn	numr
1	11	64	12.00	100	6.45×10^-4	31024	1000	3142	42932
2	12	64	12.50	100	2.89×10^-4	69216	1500	6342	87732
3	12.2	64	12.46	100	1.43×10^-4	140112	2000	11998	166916
4	12.5	64	12.82	100	1.09×10^-4	183792	3000	15998	222916
5	14.5	64	14.83	100	7.70×10^-5	259672	4500	22998	320916
6	25	64	25.30	100	5.51×10^-5	362992	6000	31998	446916
7	0.24	64	0.30	100	2.89×10^-4	69202	200	5128	70736
8	5	64	5.23	100	2.89×10^-4	69199	800	5662	78212
9	50	64	50.76	100	2.89×10^-4	69213	2000	6840	94704
10	150	64	151.70	100	2.89×10^-4	69223	3000	7856	108928
11	600	64	604.00	100	2.89×10^-4	69142	4500	10000	138944
12	1600	64	1607.35	100	2.89×10^-4	69190	6000	12598	175316

图3 PE热解Pearson相关系数显著性检验p值图（1表示p<0.05；0表示p≥0.05）

Fig.3 p-value chart of significance test of Pearson correlation coefficient of PE-biomass co-pyrolysis (1 means p<0.05, 0 means p≥0.05)

图4 PE热解Pearson相关系数

Fig.4 Pearson correlation coefficient of PE pyrolysis

图5 PE热解机器学习模型的泛化性能比较

Fig.5 Comparison of the generalization performance machine learning models for PE pyrolysis

表2 机器学习方法轻油产物收率预测性能评估

Table 2 Performance evaluation of machine learning models for light oil product yield prediction

模型	训练集				测试集
模型	数据量	R²	MAE	RMSE	数据量	R²	MAE	RMSE
KNN	28800	1.0000	0.0025	0.0605	7200	1.0000	0.0012	0.0065
RF	28800	0.9997	0.0898	0.2101	7200	0.9997	0.0972	0.1918
ANN	28800	0.9993	0.1172	0.3426	7200	0.9995	0.1250	0.2426
MLP	28800	0.9975	0.4786	0.6282	7200	0.9974	0.4586	0.5458
XGB	28800	0.9967	0.3878	0.7228	7200	0.9930	0.5038	0.8921
GBDT	28800	0.9896	0.6642	1.2810	7200	0.9789	0.7746	1.5457
DT	28800	0.9832	0.6344	1.6254	7200	0.9775	0.6009	1.5949
LGB	28800	0.9846	0.8916	1.5563	7200	0.9733	1.0540	1.7362
SVR	28800	0.8597	1.3971	4.6972	7200	0.7283	1.9827	5.5417
LR	28800	0.4849	6.9920	9.0008	7200	0.3077	6.4794	8.8453

图6 四种模型（KNN、RF、GBDT、XGB）质量收率（YW）拟合效果

Fig.6 Effectiveness of yield fitting of four models (KNN, RF, GBDT, XGB)

图7 XGB模型收率变量特征重要性分析

Fig.7 Importance analysis of XGB model variable characteristics

图8 RF模型反应时间-收率依存关系

Fig.8 The partial dependence plot of time for yield correlation prediction by RF model

图9 RF模型反应温度-收率依存关系

Fig.9 The partial dependence of temperature for yield correlation prediction by RF model

图10 RF模型k2-气体收率依存关系

Fig.10 The partial dependence of k2 for gas yield correlation prediction by RF model

图11 KNN模型轻油产物收率预测

Fig.11 Prediction of light oil yield by KNN model

参考文献 36

16	Westerhout R W J, Kuipers J A M, van Swaaij W P M. Experimental determination of the yield of pyrolysis products of polyethene and polypropene. Influence of reaction conditions[J]. Industrial & Engineering Chemistry Research, 1998, 37(3): 841-847.
17	Onwudili J A, Insura N, Williams P T. Composition of products from the pyrolysis of polyethylene and polystyrene in a closed batch reactor: effects of temperature and residence time[J]. Journal of Analytical and Applied Pyrolysis, 2009, 86(2): 293-303.
18	Chen D Z, Yin L J, Wang H, et al. Pyrolysis technologies for municipal solid waste: a review[J]. Waste Management, 2014, 34(12): 2466-2486.
19	Bagri R, Williams P T. Catalytic pyrolysis of polyethylene[J]. Journal of Analytical and Applied Pyrolysis, 2002, 63(1): 29-41.
20	Ratnasari D K, Nahil M A, Williams P T. Catalytic pyrolysis of waste plastics using staged catalysis for production of gasoline range hydrocarbon oils[J]. Journal of Analytical and Applied Pyrolysis, 2017, 124: 631-637.
21	Achilias D S, Roupakias C, Megalokonomos P, et al. Chemical recycling of plastic wastes made from polyethylene (LDPE and HDPE) and polypropylene (PP)[J]. Journal of Hazardous Materials, 2007, 149(3): 536-542.
22	张雪, 白雪峰, 赵明. 废塑料热解特性研究[J]. 化学与粘合, 2015, 37(2): 107-110.
	Zhang X, Bai X F, Zhao M. Study on pyrolysis characteristics of waste plastics[J]. Chemistry and Adhesion, 2015, 37(2): 107-110.
23	Zhao D T, Wang X H, Miller J B, et al. The chemistry and kinetics of polyethylene pyrolysis: a process to produce fuels and chemicals[J]. ChemSusChem, 2020, 13(7): 1764-1774.
24	Li J, Yu D, Pan L J, et al. Recent advances in plastic waste pyrolysis for liquid fuel production: critical factors and machine learning applications[J]. Applied Energy, 2023, 346: 121350.
25	Prasertpong P, Onsree T, Khuenkaeo N, et al. Exposing and understanding synergistic effects in co-pyrolysis of biomass and plastic waste via machine learning[J]. Bioresource Technology, 2023, 369: 128419.
26	Belden E R, Rando M, Ferrara O G, et al. Machine learning predictions of oil yields obtained by plastic pyrolysis and application to thermodynamic analysis[J]. ACS Engineering Au, 2022, 3(2): 91-101.
1	Meijer L J J, van Emmerik T, van der Ent R, et al. More than 1000 rivers account for 80% of global riverine plastic emissions into the ocean[J]. Science Advances, 2021, 7(18): eaaz5803.
2	Nielsen T D, Hasselbalch J, Holmberg K, et al. Politics and the plastic crisis: a review throughout the plastic life cycle[J]. Wiley Interdisciplinary Reviews: Energy and Environment, 2020, 9(1): e360.
3	Bahij S, Omary S, Feugeas F, et al. Fresh and hardened properties of concrete containing different forms of plastic waste—a review[J]. Waste Management, 2020, 113: 157-175.
4	Geyer R, Jambeck J R, Law K L. Production, use, and fate of all plastics ever made[J]. Science Advances, 2017, 3(7): e1700782.
5	Groh K J, Backhaus T, Carney-Almroth B, et al. Overview of known plastic packaging-associated chemicals and their hazards[J]. Science of the Total Environment, 2019, 651: 3253-3268.
6	Al-Salem S M, Lettieri P, Baeyens J. Recycling and recovery routes of plastic solid waste (PSW): a review[J]. Waste Management, 2009, 29(10): 2625-2643.
7	van Geem K M. Plastic waste recycling is gaining momentum[J]. Science, 2023, 381(6658): 607-608.
8	Rajesh Banu J, Godvin Sharmila V. A systematic review on plastic waste conversion for a circular economy: recent trends and emerging technologies[J]. Catalysis Science & Technology, 2023, 13(8): 2291-2302.
9	胡延庆, 胡凡, 周剑池, 等. 废弃塑料回收与转化的研究进展[J]. 中国塑料, 2024, 38(4): 79-87.
	Hu Y Q, Hu F, Zhou J C, et al. Research progress in upcycling of waste plastics[J]. China Plastics, 2024, 38(4): 79-87.
10	Anuar Sharuddin S D, Abnisa F, Wan Daud W M A, et al. A review on pyrolysis of plastic wastes[J]. Energy Conversion and Management, 2016, 115: 308-326.
11	Maqsood T, Dai J Z, Zhang Y N, et al. Pyrolysis of plastic species: a review of resources and products[J]. Journal of Analytical and Applied Pyrolysis, 2021, 159: 105295.
12	Dave P N, Joshi A K. Plasma pyrolysis and gasification of plastics waste—a review[J]. Journal of Scientific & Industrial Research, 2010, 69(3): 177-179.
13	Solowski G, Shalaby M, Özdemir F A. Plastic and waste tire pyrolysis focused on hydrogen production—a review[J]. Hydrogen, 2022, 3(4): 531-549.
14	孙艺蕾, 马跃, 李术元, 等. 聚烯烃塑料的热解和催化热解研究进展[J]. 化工进展, 2021, 40(5): 2784-2801.
	Sun Y L, Ma Y, Li S Y, et al. Research progress in the pyrolysis and catalytic pyrolysis of waste polyolefin plastics[J]. Chemical Industry and Engineering Progress, 2021, 40(5): 2784-2801.
15	Singh S, Patil T, Tekade S P, et al. Studies on individual pyrolysis and co-pyrolysis of corn cob and polyethylene: thermal degradation behavior, possible synergism, kinetics, and thermodynamic analysis[J]. Science of the Total Environment, 2021, 783: 147004.
27	Cheng Y, Ekici E, Yildiz G, et al. Applied machine learning for prediction of waste plastic pyrolysis towards valuable fuel and chemicals production[J]. Journal of Analytical and Applied Pyrolysis, 2023, 169: 105857.
28	Armenise S, Wong S, Ramírez-Velásquez J M, et al. Application of computational approach in plastic pyrolysis kinetic modelling: a review[J]. Reaction Kinetics, Mechanisms and Catalysis, 2021, 134(2): 591-614.
29	Hua F, Fu Z, Yang S Q, et al. Simulating polyethylene pyrolysis from a generalized molecular-level kinetic model[J]. Journal of Analytical and Applied Pyrolysis, 2023, 170: 105913.
30	Fu Z, Sun Q D, Hua F, et al. A molecular-level kinetic model for the primary and secondary reactions of polypropylene pyrolysis[J]. Journal of Analytical and Applied Pyrolysis, 2023, 175: 106182.
31	Quann R J, Jaffe S B. Structure-oriented lumping: describing the chemistry of complex hydrocarbon mixtures[J]. Industrial & Engineering Chemistry Research, 1992, 31(11): 2483-2497.
32	Ghosh P, Jaffe S B. Detailed composition-based model for predicting the cetane number of diesel fuels[J]. Industrial & Engineering Chemistry Research, 2006, 45(1): 346-351.
33	Ghosh P, Chawla B, Joshi P V, et al. Prediction of chromatographic retention times for aromatic hydrocarbons[J]. Energy & Fuels, 2006, 20(2): 609-619.
34	王睿通, 刘纪昌, 仲从伟, 等. 基于结构导向集总的催化重整分子水平反应动力学模型[J]. 石油学报(石油加工), 2020, 36(1): 95-105.
	Wang R T, Liu J C, Zhong C W, et al. Reaction kinetic model for catalytic reforming based on structure oriented lumping[J]. Acta Petrolei Sinica (Petroleum Processing Section), 2020, 36(1): 95-105.
35	Hua F, Fu Z, Cheng Y. A simplified and effective molecular-level kinetic model for plastic pyrolysis [J]. Chemical Engineering Science, 2022, 264: 118146.
36	Dubdub I, Al-Yaari M. Pyrolysis of low density polyethylene: kinetic study using TGA data and ANN prediction[J]. Polymers, 2020, 12(4): 891.

编辑推荐

Metrics

阅读次数

全文

347

HTML			PDF

最新录用	在线预览	正式出版	最新录用	在线预览	正式出版
0	0	6	176	0	165

来源	本网站	其他网站

次数	58	289
比例	17%	83%

摘要

190

最新录用	在线预览	正式出版

69	0	121

来源	本网站	其他网站

次数	28	163
比例	15%	85%

[1]	董新宇, 边龙飞, 杨怡怡, 张宇轩, 刘璐, 王腾. 冷却条件下倾斜上升管S-CO₂流动与传热特性研究[J]. 化工学报, 2024, 75(S1): 195-205.
[2]	郭骐瑞, 任丽媛, 陈康, 黄翔宇, 马卫华, 肖乐勤, 周伟良. 用于HTPB推进剂浆料的静态混合管数值模拟[J]. 化工学报, 2024, 75(S1): 206-216.
[3]	李匡奚, 于佩潜, 王江云, 魏浩然, 郑志刚, 冯留海. 微气泡旋流气浮装置内流动分析与结构优化[J]. 化工学报, 2024, 75(S1): 223-234.
[4]	赵昂然, 韩永强, 王志鹏, 李鹏飞, 许亚伟, 佟会玲. 常温条件下赤泥同时脱硫脱硝实验研究[J]. 化工学报, 2024, 75(S1): 276-282.
[5]	徐英宇, 杨国强, 彭璟, 孙海宁, 张志炳. 微界面高级氧化处理煤化工废水的研究[J]. 化工学报, 2024, 75(S1): 283-291.
[6]	王新月, 徐小虎, 张海洋, 尹春华. 维生素A醋酸酯/环糊精包合及性质研究[J]. 化工学报, 2024, 75(S1): 321-328.
[7]	汪张洲, 唐天琪, 夏嘉俊, 何玉荣. 基于复合相变材料的电池热管理性能模拟[J]. 化工学报, 2024, 75(S1): 329-338.
[8]	胡俭, 姜静华, 范生军, 刘建浩, 邹海江, 蔡皖龙, 王沣浩. 中深层U型地埋管换热器取热特性研究[J]. 化工学报, 2024, 75(S1): 76-84.
[9]	钟屹, 周仕遇, 纠连朝, 李钰晓, 吴豪江, 周智勇. 废旧磷酸铁锂电池正极材料直接修复再生研究进展[J]. 化工学报, 2024, 75(S1): 1-13.
[10]	任冠宇, 张义飞, 李新泽, 杜文静. 翼型印刷电路板式换热器流动传热特性数值研究[J]. 化工学报, 2024, 75(S1): 108-117.
[11]	杨勇, 祖子轩, 李煜坤, 王东亮, 范宗良, 周怀荣. T型圆柱形微通道内CO₂碱液吸收数值模拟[J]. 化工学报, 2024, 75(S1): 135-142.
[12]	黄俊豪, 庞克亮, 孙方远, 刘福军, 谷致远, 韩龙, 段衍泉, 冯妍卉. 干熄炉料钟结构对焦炭布料粒径均匀度影响的模拟研究[J]. 化工学报, 2024, 75(S1): 158-169.
[13]	吴学红, 韦新, 侯加文, 吕财, 刘勇, 刘鹤, 常志娟. 热解法制备碳纳米管及其在散热涂层中的应用研究[J]. 化工学报, 2024, 75(9): 3360-3368.
[14]	丁湧, 李文建, 陈昭宇, 曹立辉, 刘轩铭, 任强强, 胡松, 向军. 废旧晶体硅光伏组件EVA有氧热解动力学与产物特性[J]. 化工学报, 2024, 75(9): 3310-3319.
[15]	陈巨辉, 苏潼, 李丹, 陈立伟, 吕文生, 孟凡奇. 翅形扰流片作用下的微通道换热特性[J]. 化工学报, 2024, 75(9): 3122-3132.