現今電子產品的功能進步快速，且產品的生命週期短暫，製造業為了快速回應技術的發展，引進AI的技術發展智慧製造。PCB經過回焊爐內不同區間的溫度設定，而形成的溫度曲線是判斷PCB品質好壞的重要指標之一。本研究主要使用雙向長短期記憶神經網路模型預測PCB生產的溫度曲線，並利用堆疊自編碼器進行特徵降維，提升回焊爐的預測績效。為了找出最佳的堆疊自編碼器組合，本研究使用基因演算法得知使用2層的堆疊以及先後從16個特徵降維成15維，再降維成12維作為預測模型的輸入資料。本研究提出結合堆疊自編碼器於雙向長短期記憶神經網路模型的技術，應用於回焊爐生產的預測及品質的調控，進而協助工程師設定生產作業之參數，降低工廠生產成本。
由於本研究瑕疵數據不足，因此透過擴增數據訓練加強對於瑕疵樣本的判斷。根據實驗結果，本研究所提出的方法比起一般模型、特徵篩選與單層自編碼器的特徵降維有較佳的效能，而透過假設檢定，預測曲線的誤差比起一般模型有顯著的下降，也有信心證明型二錯誤較一般模型的結果低。從實驗結果證明，透過適當的資料擴增，並且加入堆疊自編碼器於雙向長短期記憶神經網路模型，在預測回焊爐的溫度曲線以及後續的品質判斷，有較好的表現。

At present, the functions of electronic products are advancing rapidly and the product life cycle is short. In order to respond quickly to the development of technology, the manufacturing industry has applied AI technology to develop smart manufacturing. The temperature profile formed by heating the PCB in the different zones of the reflow oven is an essential factor in judging the PCB's quality. This study focuses on predicting the temperature profile of PCB production using a Bidirectional Long Short-Term Memory (Bi-LSTM) Neural Network model and using a Stacked Autoencoder (SAE) for feature reduction to improve the prediction performance. To find the best combination of SAE, this study uses the genetic algorithm to know using a 2-layer stacking and reducing the dimensionality from 16 features to 15 dimensions and then to 12 dimensions as input data for the prediction model. This study proposes combining SAE in a Bi-LSTM for prediction and quality control of reflow oven production and then to help engineers define the production operation's parameters and reduce the factory's production cost.
Since the insufficient defective data in this study, training in augmented data enhances the judgment of defective samples. According to the experimental results, the proposed method performs better than the general model, feature selection and feature reduction of the single-layer autoencoder. Through hypothesis testing, the error of the prediction profile is significantly lower than that of the general model, and we are also confident that the Type II error is lower than that of the general model. Experimental results indicated that appropriately augmenting the data and integrating the SAE into a Bi-LSTM performs better in predicting the reflow oven temperature profile and quality judgment.

Adem, K., Kiliçarslan, S., & Cömert, O. (2019). Classification and diagnosis of cervical cancer with stacked autoencoder and softmax classification. Expert Systems with Applications, 115, 557-564.
Alelaumi, S., Khader, N., He, J., Lam, S., & Yoon, S. W. (2021). Residue buildup predictive modeling for stencil cleaning profile decision-making using recurrent neural network. Robotics and Computer-Integrated Manufacturing, 68, 102041.
An, J., & Cho, S. (2015). Variational autoencoder based anomaly detection using reconstruction probability. Special Lecture on IE, 2(1), 1-18.
Anowar, F., Sadaoui, S., & Selim, B. (2021). Conceptual and empirical comparison of dimensionality reduction algorithms (pca, kpca, lda, mds, svd, lle, isomap, le, ica, t-sne). Computer Science Review, 40, 100378.
Box, G. E., Jenkins, G. M., Reinsel, G. C., & Ljung, G. M. (2015). Time series analysis: forecasting and control. John Wiley & Sons.
Chang, Y.-M., Wei, C.-C., Chen, J., & Hsieh, P. (2019). An implementation of health prediction in SMT solder joint via machine learning. 2019 ieee international conference on big data and smart computing (bigcomp),
Che, D., Liu, Q., Rasheed, K., & Tao, X. (2011). Decision tree and ensemble learning algorithms with their applications in bioinformatics. Software tools and algorithms for biological systems, 191-199.
Chizi, B., & Maimon, O. (2009). Dimension reduction and feature selection. In Data mining and knowledge discovery handbook (pp. 83-100). Springer.
Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014). Empirical evaluation of gated recurrent neural networks on sequence modeling. arXiv preprint arXiv:1412.3555.
Conway, P. P., Whalley, D. C., Wilkinson, M., & Williams, D. J. (1997). Automated adaptive control of the reflow soldering of electronic assemblies. Twenty First IEEE/CPMT International Electronics Manufacturing Technology Symposium Proceedings 1997 IEMT Symposium,
Deb, K., & Agrawal, R. B. (1995). Simulated binary crossover for continuous search space. Complex systems, 9(2), 115-148.
Deb, K., & Goyal, M. (1996). A combined genetic adaptive search (GeneAS) for engineering design. Computer Science and informatics, 26, 30-45.
Deb, K., Pratap, A., Agarwal, S., & Meyarivan, T. (2002). A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE transactions on evolutionary computation, 6(2), 182-197.
Devassy, B. M., & George, S. (2020). Dimensionality reduction and visualisation of hyperspectral ink data using t-SNE. Forensic science international, 311, 110194.
Gao, J., Wu, Y., & Ding, H. (2007). Optimization of a reflow soldering process based on the heating factor. Soldering & surface mount technology, 19(1), 28-33.
Gong, Y., Li, Q., & Yang, D. (2006). The optimization of reflow soldering temperature profile based on simulation. 2006 7th International Conference on Electronic Packaging Technology,
Graves, A., Mohamed, A.-r., & Hinton, G. (2013). Speech recognition with deep recurrent neural networks. 2013 IEEE international conference on acoustics, speech and signal processing,
Heaton, J. (2016). An empirical analysis of feature engineering for predictive modeling. SoutheastCon 2016,
Hinton, G. E., & Salakhutdinov, R. R. (2006). Reducing the dimensionality of data with neural networks. science, 313(5786), 504-507.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.
Jiang, R., Fu, W., Wen, L., Hao, S., & Hong, R. (2016). Dimensionality reduction on anchorgraph with an efficient locality preserving projection. Neurocomputing, 187, 109-118.
Jindal, C., Salls, C., Aghakhani, H., Long, K., Kruegel, C., & Vigna, G. (2019). Neurlux: dynamic malware analysis without feature engineering. Proceedings of the 35th Annual Computer Security Applications Conference,
Karasu, S., & Altan, A. (2019). Recognition model for solar radiation time series based on random forest with feature selection approach. 2019 11th international conference on electrical and electronics engineering (ELECO),
Khurana, U., Turaga, D., Samulowitz, H., & Parthasrathy, S. (2016). Cognito: Automated feature engineering for supervised learning. 2016 IEEE 16th International Conference on Data Mining Workshops (ICDMW),
Kim, H., Drake, B. L., & Park, H. (2007). Multiclass classifiers based on dimension reduction with generalized LDA. Pattern recognition, 40(11), 2939-2945.
Kwak, N., Kim, C., & Kim, H. (2008). Dimensionality reduction based on ICA for regression problems. Neurocomputing, 71(13-15), 2596-2603.
Lai, Y., Kataoka, J., Pan, K., Ha, J., Yang, J., Deo, K. A., Xu, J., Yin, P., Cai, C., & Park, S. (2022). A Deep Learning Approach for Reflow Profile Prediction. 2022 IEEE 72nd Electronic Components and Technology Conference (ECTC),
Li, C.-q., & Wu, Z.-h. (2006). SPC system analysis and design of reflow soldering process. 2006 7th International Conference on Electronic Packaging Technology,
Li, Y., He, J., Won, D., & Yoon, S. W. (2021). Noncontact Reflow Oven Thermal Profile Prediction Based on Artificial Neural Network. IEEE Transactions on Components, Packaging and Manufacturing Technology, 11(12), 2229-2237.
Liu, G., Bao, H., & Han, B. (2018). A stacked autoencoder-based deep neural network for achieving gearbox fault diagnosis. Mathematical Problems in Engineering, 2018.
Liu, X., & Xv, L. (2019). Abstract summarization based on the combination of transformer and lstm. 2019 International Conference on Intelligent Computing, Automation and Systems (ICICAS),
Ma, J., & Yuan, Y. (2019). Dimension reduction of image deep feature using PCA. Journal of Visual Communication and Image Representation, 63, 102578.
Makhzani, A., Shlens, J., Jaitly, N., Goodfellow, I., & Frey, B. (2015). Adversarial autoencoders. arXiv preprint arXiv:1511.05644.
Mariet, Z., & Kuznetsov, V. (2019). Foundations of sequence-to-sequence modeling for time series. The 22nd international conference on artificial intelligence and statistics,
Mousavi, S. M., Ellsworth, W. L., Zhu, W., Chuang, L. Y., & Beroza, G. C. (2020). Earthquake transformer—an attentive deep-learning model for simultaneous earthquake detection and phase picking. Nature communications, 11(1), 1-12.
Nti, K. O., Adekoya, A., & Weyori, B. (2019). Random forest based feature selection of macroeconomic variables for stock market prediction. American Journal of Applied Sciences, 16(7), 200.212.
Pan, Y., Ge, S. S., & Al Mamun, A. (2009). Weighted locally linear embedding for dimension reduction. Pattern recognition, 42(5), 798-811.
Saraswati, A., Hagenbuchner, M., & Tsoi, A. C. (2018). High-resolution Self-Organizing Maps for advanced visualization and dimension reduction. Neural Networks, 105, 166-184.
Shahid, F., Zameer, A., & Muneeb, M. (2020). Predictions for COVID-19 with deep learning models of LSTM, GRU and Bi-LSTM. Chaos, Solitons & Fractals, 140, 110212.
Shewalkar, A. N. (2018). Comparison of rnn, lstm and gru on speech recognition data.
Siami-Namini, S., Tavakoli, N., & Namin, A. S. (2019). The performance of LSTM and BiLSTM in forecasting time series. 2019 IEEE International Conference on Big Data (Big Data),
Singh, S., & Mahmood, A. (2021). The NLP cookbook: modern recipes for transformer based deep learning architectures. IEEE Access, 9, 68675-68702.
Su, Y., Srihari, K., & Emerson, C. (1997). A profile identification system for surface mount printed circuit board assembly. Computers & industrial engineering, 33(1-2), 377-380.
Sutskever, I., Vinyals, O., & Le, Q. V. (2014). Sequence to sequence learning with neural networks. Advances in neural information processing systems, 27.
Tsai, T.-N. (2012). Thermal parameters optimization of a reflow soldering profile in printed circuit board assembly: A comparative study. Applied Soft Computing, 12(8), 2601-2613.
Tsai, T. N., & Yang, T. (2005). A neuro‐computing approach to the thermal profile control of the second‐side reflow process in surface mount assembly. Journal of Manufacturing Technology Management.
Tu, S. T., Chen, J. Y., Yang, W., & Sun, H. (2011). Laplacian eigenmaps-based polarimetric dimensionality reduction for SAR image classification. IEEE Transactions on Geoscience and Remote Sensing, 50(1), 170-179.
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., & Polosukhin, I. (2017). Attention is all you need. Advances in neural information processing systems, 30.
Vincent, P., Larochelle, H., Bengio, Y., & Manzagol, P.-A. (2008). Extracting and composing robust features with denoising autoencoders. Proceedings of the 25th international conference on Machine learning,
Wang, Z., Li, W., Tong, H., & Xu, Y. (2006). A novel on-board temperature monitoring approach in the reflow soldering process. 2006 IEEE/RSJ International Conference on Intelligent Robots and Systems,
Whalley, D. C. (2002). A simplified model of the reflow soldering process. ITherm 2002. Eighth Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (Cat. No. 02CH37258),
Wu, C.-H., Srihari, K., & McLenaghan, A. (1996). A knowledge-based thermal profile identification advisor for surface mount PCB assembly. The International Journal of Advanced Manufacturing Technology, 11(5), 343-352.
Yan, Y., Qi, L., Wang, J., Lin, Y., & Chen, L. (2020). A network intrusion detection method based on stacked autoencoder and LSTM. ICC 2020-2020 IEEE International Conference on Communications (ICC),
Yang, J., & Honavar, V. (1998). Feature subset selection using a genetic algorithm. In Feature extraction, construction and selection (pp. 117-136). Springer.
Yang, Y., Wu, Q. J., & Wang, Y. (2016). Autoencoder with invertible functions for dimension reduction and image reconstruction. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 48(7), 1065-1079.
Zabalza, J., Ren, J., Zheng, J., Zhao, H., Qing, C., Yang, Z., Du, P., & Marshall, S. (2016). Novel segmented stacked autoencoder for effective dimensionality reduction and feature extraction in hyperspectral imaging. Neurocomputing, 185, 1-10.
Zhang, G. P. (2003). Time series forecasting using a hybrid ARIMA and neural network model. Neurocomputing, 50, 159-175.
Zou, H., & Hastie, T. (2005). Regularization and variable selection via the elastic net. Journal of the royal statistical society: series B (statistical methodology), 67(2), 301-320.