本研究運用人工智慧於流化床氣化爐，探討廢棄香菇太空包與聚乙烯（PE）塑膠共氣化產氣的預測研究。廢棄香菇包具備高灰份與較低熱值的特性，而PE則具有高熱值與零灰份的優點，但是容易受熱融化沾黏導致進料操作困難，因此共氣化能互補兩者的優缺點，提升能源轉換的潛力。實驗數據來自先前本實驗室1kWth流化床共氣化實驗，操作參數設計依據田口法，資料包含氣化溫度、原料混摻比例、催化劑含量與氣體產物等資訊。本研究首先以傅立葉分析濾除高頻雜訊後重建疏相區溫度（T1）以避免資料遺失造成模型性能下降。在人工智慧模型部分，採用多項式回歸、隨機森林（RF）、XGBoost、LSTM、GRU與多層感知機（MLP）等模型，並透過Optuna進行超參數優化。研究結果顯示，隨機森林與多項式回歸皆能有效重建T1資料，並顯著提升產氣預測準確率。在產氣預測方面，多數模型預測結果的R²值高達0.9以上，其中RF表現最為穩定。此外，本研究亦針對MLP模型進行效能提升，導入Boosting與Stacking技術進行優化。藉由集成多個弱學習器的預測結果來修正MLP模型的偏差，提升整體預測表現。結果顯示，透過Boosting後，MLP模型在CO預測的決定係數R²由原始約0.849提升至0.866，而使用Stacking MLP模型的決定係數R²由原始約0.849提升至0.874，顯示集成法有助於進一步提升模型效能，使模型更加可靠。本研究針對CO、H₂、CH₄與CO₂四種氣體進行時序預測，結果顯示XGBoost與LSTM在R²指標上整體優於GRU，於短歷史長度下皆達0.9以上，其中LSTM略勝一籌。長歷史設定中，XGBoost於CH₄與CO₂的「長記憶（30s）+中未來（60s）」情境表現突出。GRU在中長記憶條件下MAE偏高，可能受限於資料量與泛化能力，但訓練效率較佳。整體訓練時間排序為XGBoost、GRU、LSTM。XGBoost與LSTM於各歷史設定下MAE多維持1以下，具穩定預測能力；LSTM於RMSE表現優勢明顯，顯示其在處理高時間依賴性資料時具備更佳誤差控制能力。

This study investigates the application of artificial intelligence techniques to predict syngas production from the co-gasification of spent mushroom substrate (SMS) and polyethylene (PE) plastic waste in a 1 kWth fluidized bed gasifier. SMS and PE exhibit complementary properties in heating value and ash content. Operating parameters were designed using the Taguchi method, and Fourier analysis was applied to reconstruct coarse temperature (T1) signals. Machine learning models, including polynomial regression (PR), random forest (RF), and multilayer perceptron (MLP), were developed with hyperparameters optimized via Optuna. PR and RF achieved R² values above 0.9 in both T1 reconstruction and syngas prediction. For CO prediction, MLP performance improved from an R² of 0.849 to 0.874 through boosting and stacking. In time-series forecasting, XGBoost and LSTM outperformed GRU under short-history conditions, with LSTM showing the lowest RMSE and highest accuracy. These results highlight the effectiveness of integrating experimental methods with machine learning and ensemble techniques to enhance prediction performance in biomass-plastic co-gasification systems.

[1] Climate Change:Atmospheric Carbon Dioxide.https://www.climate.gov/news features/understanding-climate/climate-change-atmospheric-carbon-dioxide.
[2] International Energy Agency,“Climate Change – Topics,” IEA. [Online]. Available: https://www.iea.org/topics/climate-change
[3] Steve Sorrell. Reducing energy demand: A review of issues, challenges and approaches. Renewable and Sustainable Energy Reviews, 47:74–82, July 2015.
[4] E. Monteiro and S. Ferreira, “Some perspectives for the gasification process in the energy transition context,” *Cleaner Production Letters*, vol. 5, 100041, Nov. 2023. [Online].
[5] N. Khan, E. Kalair, N. Abas, A.R. Kalair, and A. Kalair. Energy transition from molecules to atoms and photons. Engineering Science and Technology, an International Journal, 22(1):185–214, February 2019.
[6] S. Harichandan, S. K. Kar, R. Bansal, S. K. Mishra, M. S. Balathanigaimani, and M. Dash, “Energy transition research: A bibliometric mapping of current findings and direction for future research,” *Cleaner Production Letters*, vol. 3, 100026, Nov. 2022.
[7] International Energy Agency, *Net Zero by 2050: A Roadmap for the Global Energy Sector*, Paris: IEA, Oct. 2021.
[8] 經濟部, *2020年能源政策白皮書*,台北市:經濟部能源局,2020年11月. [線上]. Available: https://www.moeaboe.gov.tw/ECW/populace/content/ContentLink.aspx?menu_id=1573.
[9] 經濟部能源署, *112年度全國電力資源供需報告*, 台北市: 經濟部能源署, 2023年12月. [線上]. Available: https://www.moeaea.gov.tw/
[10] Antonio Tursi. A review on biomass: importance, chemistry, classification, and conversion. Biofuel Research Journal, 6(2):962–979, June 2019.
[11] Chung-Yu Chang and Guan-Bang Chen. Modeling of biomass gasification in fluidized bed. December 2022.
[12] J. T. Chen, H. D. Shih, Y. N. Chen, and P. J. Ann, “Mushroom industry policies and guidance measures,” Proceedings of the Symposium on Mushroom Industry Development, Taiwan Agricultural Research Institute, Council of Agriculture,.
[13] E. Atallah, J. Zeaiter, M. N. Ahmad, J. J. Leahy, and W. Kwapinski, “Hydrothermal carbonization of spent mushroom compost waste compared against torrefaction and pyrolysis,” Fuel Process. Technol., vol. 216, p. 106795, Jun. 2021, doi: 10.1016/j.fuproc.2021.106795.
[14] Y.-C. Chang, Y.-T. Chen, and Y.-S. Lin, “Design and development of mushroom classification system based on deep learning,” in Proc. Symp. Mushroom Industry Development, Taichung, Taiwan, Oct. 2011, pp. 49–56. (in Chinese)
[15] Statista, “Global plastics production: share of various countries and regions 2023,” Statista, Jul. 2024
[16] Roland Geyer, Jenna R. Jambeck, and Kara Lavender Law. Production, use, and fate of all plastics ever made. Science Advances, 3(7):e1700782, July 2017.
[17] Sahar Safarian, Rúnar Unnþórsson, and Christiaan Richter. A review of biomass gasification modelling. Renewable and Sustainable Energy Reviews, 110:378–391, August 2019.
[18] Muhammad Saghir, Mohammad Rehan, and Abdul-Sattar Nizami. Recent trends in gasification based waste-to-energy. Gasification for low-grade feedstock, 2018:97–113, 2018.
[19] A. Molino, F. Nanna, Y. Ding, A. Bikson, and G. Braccio, “Biomass gasification technology: The state of the art overview,” J. Energy Chem., vol. 25, no. 1, pp. 10–25, Jan. 2016, doi: 10.1016/j.jechem.2015.11.005.
[20] Fatma Alfarra, H. Kurtulus Ozcan, Pınar Cihan, Atakan Ongen, Senem Yazici Guvenc, and Mirac Nur Ciner. Alfarra et al. - 2024 - Artificial intelligence methods for modeling gasification. Environmental Monitoring and Assessment, 196(3):309, February 2024.
[21] Carine T. Alves, Jude A. Onwudili, Payam Ghorbannezhad, and Shogo Kumagai. A review of the thermochemistries of biomass gasification and utilisation of gas products. Sustainable Energy & Fuels, 7(15):3505–3540, 2023.
[22] Great C. Umenweke, Inioluwa Christianah Afolabi, Emmanuel I. Epelle, and Jude A. Okolie. Machine learning methods for modeling conventional and hydrothermal gasification of waste biomass A review. Bioresource Technology Reports, 17:100976, February 2022.
[23] Mortaza Aghbashlo, Wanxi Peng, Meisam Tabatabaei, Soteris A. Kalogirou, Salman Soltanian, Homa Hosseinzadeh-Bandbafha, Omid Mahian, and Su Shiung Lam. Machine learning technology in biodiesel research: A review. Progress in Energy and Combustion Science, 85:100904, July 2021.
[24] Maria Puig-Arnavat, J. Alfredo Hernández, Joan Carles Bruno, and Alberto Coronas. Artificial neural network models for biomass gasification in fluidized bed gasifiers. Biomass and Bioenergy, 49:279–289, February 2013.
[25] Anh Tuan Hoang, ZuoHua Huang, Sandro Nižetić, Ashok Pandey, Xuan Phuong Nguyen, Rafael Luque, Hwai Chyuan Ong, Zafar Said, Tri Hieu Le, and Van Viet Pham. Characteristics of hydrogen production from steam gasification of plant-originated lignocellulosic biomass and its prospects in Vietnam. International Journal of Hydrogen Energy, 47(7):4394–4425, January 2022.
[26] V. S. Sikarwar et al., “An overview of advances in biomass gasification,” Energy Environ. Sci., vol. 9, no. 10, pp. 2939–2977, 2016, doi: 10.1039/c6ee00935b.
[27] A. Erkiaga, G. Lopez, M. Amutio, J. Bilbao, and M. Olazar, “Influence of operating conditions on the steam gasification of biomass in a conical spouted bed reactor,” Chem. Eng. J., vol. 237, pp. 259–267, Feb. 2014, doi: 10.1016/j.cej.2013.10.018.
[28] M. Cortazar et al., “Fe/olivine as primary catalyst in the biomass steam gasification in a fountain confined spouted bed reactor,” J. Ind. Eng. Chem., vol. 99, pp. 364–379, Jul. 2021, doi: 10.1016/j.jiec.2021.04.046.
[29] N. M. Nguyen, F. Alobaid, J. May, J. Peters, and B. Epple, “Experimental study on steam gasification of torrefied woodchips in a bubbling fluidized bed reactor,” Energy, vol. 202, p. 117744, Jul. 2020, doi: 10.1016/j.energy.2020.117744.
[30] S. Sharma and P. N. Sheth, “Air–steam biomass gasification: Experiments, modeling and simulation,” Energy Convers. Manag., vol. 110, pp. 307–318, Feb. 2016, doi: 10.1016/j.enconman.2015.12.030.
[31] I. Narváez, A. Orío, M. P. Aznar, and J. Corella, “Biomass Gasification with Air in an Atmospheric Bubbling Fluidized Bed. Effect of Six Operational Variables on the Quality of the Produced Raw Gas,” Ind. Eng. Chem. Res., vol. 35, no. 7, pp. 2110–2120, Jan. 1996, doi: 10.1021/ie9507540.
[32] S. Luo, B. Xiao, Z. Hu, S. Liu, X. Guo, and M. He, “Hydrogen-rich gas from catalytic steam gasification of biomass in a fixed bed reactor: Influence of temperature and steam on gasification performance,” Int. J. Hydrog. Energy, vol. 34, no. 5, pp. 2191–2194, Mar. 2009, doi: 10.1016/j.ijhydene.2008.12.075.
[33] Q Zhu, J M Jones, A Williams, and K M Thomas. The predictions of coal/char combustion rate using an artificial neural network approach.
[34] Sheng Su and Juan Wang. Machine learning prediction of contents of oxygenated components in bio-oil using extreme gradient boosting method under different pyrolysis conditions. Bioresource Technology, 379:129040, July 2023.
[35] Seyed Hashem Samadi, Barat Ghobadian, Mohsen Nosrati, and Mahdi Rezaei.  Investigation of factors affecting performance of a downdraft fixed bed gasifier using optimized MLP neural networks approach. Fuel, 333:126249, February 2023.
[36] Imtiaz Ali Jamro, Abdul Raheem, Salim Khoso, Humair Ahmed Baloch, Akash Kumar, Guanyi Chen, Waheed Ali Bhagat, Terrence Wenga, and Wenchao Ma. Investigation of enhanced H₂ production from municipal solid waste gasification via artificial neural network with data on tar compounds. Journal of Environmental Management, 328:117014, February 2023.
[37] Jun Young Kim, Dongjae Kim, Zezhong John Li, Claudio Dariva, Yankai Cao, and Naoko Ellis. Predicting and optimizing syngas production from fluidized bed biomassgasifiers: A machine learning approach. Energy, 263:125900, January 2023.
[38] Furkan Elmaz, Özgün Yücel, and Ali Yener Mutlu. Predictive modeling of biomass gasification with machine learning-based regression methods. Energy, 191:116541, January 2020.
[39] Bamidele Victor Ayodele, Siti Indati Mustapa, Ramesh Kanthasamy, Mohammed Zwawi, and Chin Kui Cheng. Modeling the prediction of hydrogen production by co-gasification of plastic and rubber wastes using machine learning algorithms. International Journal of Energy Research, 45(6):9580–9594, May 2021.
[40] Jie Li, Manu Suvarna, Lanjia Pan, Yingru Zhao, and Xiaonan Wang. A hybrid data-driven and mechanistic modelling approach for hydrothermal gasification. Applied Energy, 304:117674, December 2021.
[41] Uğur Özveren, Furkan Kartal, Senem Sezer, and Z. Sibel Özdoğan. Investigation of steam gasification in thermogravimetric analysis by means of evolved gas analysis and machine learning. Energy, 239:122232, January 2022.
[42] Ibtihaj Khurram Faridi, Evangelos Tsotsas, Wolfram Heineken, Marcus Koegler, and Abdolreza Kharaghani. Spatio-temporal prediction of temperature in fluidized bed biomass gasifier using dynamic recurrent neural network method. Applied Thermal Engineering, 219:119334, January 2023.
[43] Model Variance: Everything You Need to Know When Assessing Model Variance Skills. https://www.alooba.com/skills/concepts/machine-learning/model-variance/.
[44] Cross-Validation an overview | ScienceDirect Topics https://www.sciencedirect.com/topics/medicine-and-dentistry/cross-validation.
[45] Learning  Curve an overview |ScienceDirect Topics https://www.sciencedirect.com/topics/earth-and-planetary-sciences/learning-curve.
[46] Aurélien Géron. Hands-on Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems. Data Science / Machine Learning. O’Reilly, Beijing Boston Farnham Sebastopol Tokyo, third edition edition, 2023.
[47] Artificial Neural Network an overview | ScienceDirect Topics https://www.sciencedirect.com/topics/neuroscience/artificial-neural-network.
[48] Hung-Tai Lung, Ta-Hui Lin, Fang-Hsien Wu, and Guan-Bang Chen.  Application of Artificial Intelligence in Industrial Furnaces and Droplet Behaviors.
[49] “Multilayer Perceptron - an overview | ScienceDirect Topics.” Accessed: May 17, 2025. [Online]. Available: https://www.sciencedirect.com/topics/computer-science/multilayer-perceptron
[50] Quan Nguyen. Bayesian Optimization in Action. Manning Publications, Erscheinung sort nicht ermittelbar, 2023.
[51] Somya Mishra and Rajesh Kumar Upadhyay. Review on biomass gasification: Gasifiers, gasifying mediums, and operational parameters. Materials Science for Energy Technologies, 4:329–340, January 2021.