由於全球暖化日益嚴重，以及台灣對高效率與高可靠度能源的需求日漸增加，核能被視為一種潛在的解決方案，可有效減少碳排放，並對抗氣候變遷。然而，核能技術在應用上的安全性仍是高度關切的議題，特別是在面對異常狀況與不可預測事故風險時。因此，針對核能電廠進行故障監測與預後健康管理（PHM）成為確保其安全運轉的必要措施。
本論文旨在透過開發與評估先進的警報洪水分類（Alarm Flood Classification, AFC）方法，提升核能電廠（Nuclear Power Plants, NPP）之安全性與預測效能。為了解決現有研究中對於警報資料集缺乏的問題，本研究採用PCTRAN模擬軟體所提供之IEEE開放資料集，建立一套涵蓋多種異常情境的綜合性資料庫。此資料集能夠深入分析警報行為，並為AFC方法的效能評估提供堅實的基礎。研究中以系統化方式產生警報資料，並依據誤報率（False Alarm Rate, FAR）與漏報率（Missed Alarm Rate, MAR）之間的平衡，設置alpha參數範圍從0.00至1.00的不同警報觸發門檻值。
此外，研究進一步探討深度學習技術於此類時序資料的應用，包括循環神經網路（Recurrent Neural Network, RNN）、長短期記憶網路（Long Short-Term Memory, LSTM）、門控循環單元（Gated Recurrent Unit, GRU），以及結合滑動視窗技術之LSTM與GRU模型。透過這些模型對特徵進行時序分析與擷取，以提升分類模型的準確性與穩定性。
實驗結果顯示，深度學習技術中以GRU模型搭配滑動視窗方法進行警報分類效果最佳，當滑動視窗大小（window size）設為290，步長（step size）為1，並於最佳alpha值0.10下運作時，能達到99%的分類準確率，其加權平均的精確率（Precision）、召回率（Recall）與F1分數亦皆為99%。此結果顯示該方法具備即時反應能力與高分類效能，有助於提升核能電廠之整體運作可靠性，並在事故應對與設施營運成本方面達到最小化之效益。

Due to global warming and Taiwan’s growing need for efficient and reliable energy sources, nuclear energy is considered a potential solution for reducing carbon emissions and combating climate change. However, safety remains a critical concern due to the risks of abnormal situations and uncertain accidents. This makes monitoring and prognostic and health management (PHM) of faults essential.
The goal of this thesis is to provide the safety and efficient prognostic of nuclear power plants (NPP) by developing and evaluating advanced methods for alarm flood classification (AFC). To address a gap in existing research regarding the availability of alarm datasets, a comprehensive dataset derived from simulated data of a nuclear power plant using the IEEE open dataset from PCTRAN software was utilized. This dataset enables detailed analysis of alarm behavior and provides a foundation for evaluating AFC methods. A systematic approach was used to generate alarm data, setting thresholds based on balancing false alarm rates and miss alarm rates from alpha 0.00 to 1.00. Additionally, the research investigates the application of deep learning techniques, such as Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), LSTM and GRU with sliding window function to analyze the dataset and features that improve classification model performance.
In this thesis, I find that deep learning deals with alarm classification in GRU with a sliding window function with window size of 290 and a step size of 1 operating with an optimal alpha value of 0.10 achieving 99% accuracy, with weighted average precision, recall, and F1-score also at 99%. These contribute to enhancing overall NPP reliable operation through on-time handling, high classification accuracy, and minimized overall facility costs.

Reference
[1] Roberto Schirru, Mikael Schirru, Manoel Villas Boas Junior, Andressa dos Santos Nicolau, Alan Miranda Monteiro de Lima, Victor Henrique Cabral Pinheiro. (2020). Determination of alarm templates for decision support in nuclear power plants alarm floods using evolutionary computation, Progress in Nuclear Energy. Science Direct Journals, vol. 123. https://doi.org/10.1016/j.pnucene.2020.103308.
[2] Georgios Mitrentsis, Susanne Schmitt, David Marino, Jhelum Chakravorty, Giancarlo Dalle Ave, Faeza Hafiz, Antony Hilliard. (2024). Intelligent alarm flood management for power systems, Electric Power Systems Research. Science Direct Journals, vol. 235. https://doi.org/10.1016/j.epsr.2024.110774.
[3] Qi Ben, Xiao Xingyu, Liang Jingang, Po Li-chi Cliff, Zhang Liguo, and Tong Jiejuan. (2022). An open time-series simulated dataset covering various accidents for nuclear power plants. Scientific data, Volume 9, 766, pp. 1-14. https://doi.org/10.1038/s41597-022-01879-1.
[4] Kompella Janaki Devi, Ravichandran Karthik, Anandan, Vignesh. (2023). Application of Probabilistic Risk Assessment Approach in Nuclear Power and Space Sectors. Advances in Reliability and Safety Assessment for Critical Systems. NCRS 2022. Lecture Notes in Mechanical Engineering. Springer, Singapore, pp. 325-345. https://doi.org/10.1007/978-981-99-5049-2_25.
[5] Liu Bing, Lei Jichong, Xie Jinsen, Zhou Jianliang. (2022). Development and Validation of a Nuclear Power Plant Fault Diagnosis System Based on Deep Learning. MDPI Energies, 15, 8629, pp. 1-15. https://doi.org/10.3390/en15228629.
[6] Zhao Xingang, Kim Junyung, Warns Kyle, Wang Xinyan, Ramuhalli Pradeep, Cetiner Sacit, Kang Hyun Gook and Golay Michael. (2021). Prognostics and Health Management in Nuclear Power Plants: An Updated Method-Centric Review With Special Focus on Data Driven Methods. Frontiers Energy Research. Volume 9, 696785, pp. 1-20. https://doi.org/10.3389/fenrg.2021.696785.
[7] Agarwal Vivek, Lybeck Nancy, Pham Binh T., Rusaw Richard, and Bickford Randall. (2015). Prognostic and Health Management of Active Assets in Nuclear Power Plants. International Journal of Prognostics and Health Management, Vol.6 No.3, pp. 1-15. https://doi.org/10.36001/ijphm.2015.v6i3.2270.
[8] Hicham Lamtougui, Hicham El Moubtahij, Hassan Fouadi, Ali Yahyaouy, Khalid SATORI. (2020). Offline Arabic Handwriting Recognition Using Deep Learning: Comparative Study. 2020 IEEE Xplore International Conference on Intelligent Systems and Computer Vision (ISCV), pp.1-5, https://doi.org/10.1109/ISCV49265.2020.9204214.
[9] Peng Junjie, Jury Elizabeth C., Dönnes Pierre, Ciurtin Coziana. (2021). Machine Learning Techniques for Personalised Medicine Approaches in Immune-Mediated Chronic Inflammatory Diseases: Application and Challenges. Frontiers in pharmacology, volume 12, pp. 1-6. https://doi.org/10.3389/fphar.2021.720694.
[10] Wang Boyi and Zhang Wei. (2023). Research on Edge Network Topology Optimization Based on Machine Learning. 2023 5th International Conference on Applied Machine Learning (ICAML). Dalian, China, pp. 41-46, https://doi.org/10.1109/ICAML60083.2023.00018.
[11] Teoh Teik Toe. (2023). Convolutional neural networks for medical applications. Springer Singapore, Page. 3-5. https://doi.org/10.1007/978-981-19-8814-1
[12] Alinezhad Haniyeh Seyed, Shang Jun, and Chen Tongwen. (2022). Early Classification of Industrial Alarm Floods Based on Semisupervised Learning. IEEE Transactions on Industrial Informatics, Vol 18, No.3, pp. 1845-1853, https://doi.org/10.1109/TII.2021.3081417.
[13] Masum Shamsul, Liu Ying, and Chiverton John. (2018). Multi-step Time Series Forecasting of Electric Load Using Machine Learning Models. Springer International Publishing AG, ICAISC 2018, LNAI 10841, pp. 148-159. https://doi.org/10.1007/978-3-319-91253-0_15.
[14] Bhateja Vikrant, Carroll Fiona, Tavares João Manuel R. S., Sengar Sandeep Singh, Peer Peter. (2023). Intelligent Data Engineering and Analytics: Proceedings of the 11th International Conference on Frontiers of Intelligent Computing: Theory and Applications (FICTA 2023). Springer Nature, volume 371, pp. 26-28. https://doi.org/10.1007/978-981-99-6706-3.
[15] Zhang Yi, Kok Kiong Tan. (2004). Convergence Analysis of Recurrent Neural Networks. Springer Science Business Media Dordrecht, pp. 15-32. https://doi-org.ezproxy.lib.ncu.edu.tw/10.1007/978-1-4757-3819-3.
[16] Bianchi Maria Filippo, Maiorino Enrico, Kampffmeyer Michael C., Rizzi Antonello, and Jenssen Robert. (2017). Recurrent Neural Networks for Short-Term Load Forecasting. Springer Briefs in Computer Science, pp. 9-21. https://doi.org/10.1007/978-3-319-70338-1_2.
[17] Salem M. Fathi. (2022). Recurrent Neural Networks. Springer Nature Switzerland AG, pp.43-67. https://doi.org/10.1007/978-3-030-89929-5_3.
[18] Graves Alex. (2012). Supervised sequence labelling with recurrent neural networks. Springer-Verlag Berlin Heidelberg, SCI 385, pp. 37–45. https://doi-org.ezproxy.lib.ncu.edu.tw/10.1007/978-3-642-24797-2
[19] Chen Yakun, Li Zihao, Yang Chao, Wang Xianzhi, Long Guodong, and Xu Guandong. (2023). Adaptive Graph Recurrent Network for Multivariate Time Series Imputation. ICONIP 2022. Communications in Computer and Information Science, vol 1792, pp. 64–73. Springer, Singapore. https://doi.org/10.1007/978-981-99-1642-9_6.
[20] Dunis L. Christian, Middleton W. Peter, Theofilatos Konstantinos, and Karathanasopoulos Andreas. (2016). Modelling, Forecasting and Trading the Crack: A Sliding Window Approach to Training Neural Networks. Artificial Intelligence in Financial Markets. New Developments in Quantitative Trading and Investment. Palgrave Macmillan, London, pp.7-8. https://doi.org/10.1057/978-1-137-48880-0_3.
[21] Sucharitha B., Devi Damisetti Savitri, and Sushmitha Jakku. (2022). Image Segmentation Techniques and Optimization Algorithms for Lung Cancer Detection. Advances in Signal Processing and Communication Engineering. Lecture Notes in Electrical Engineering, vol 929, pp. 179-180. Springer, Singapore. https://doi.org/10.1007/978-981-19-5550-1_17.