本研究利用倒傳遞神經網路整合基因演算法來建立射出成型的製程參數與成品品質之間的關係，並使用基因演算法來獲取最佳製程參數，達到減少成品之體積收縮的目的。首先利用模穴內的溫度與壓力傳感器來監測實驗過程中的熔膠狀態，將數據轉換成比容的形式，然後合併非均勻比容之指標及體積收縮的指標作為整體體積收縮的指標以及實驗計畫法的反應值，並對其進行資料標準化，接著使用此數據來訓練倒傳遞神經網路模型，再將訓練完成的模型作為基因演算法的適應函數，最後分別比較使用倒傳遞神經網路整合基因演算法以及使用反應曲面法和田口法所得到的優化結果之間的差異。
本研究的結果顯示，在神經網路為5-11-1的架構下得到驗證組之平均絕對百分比誤差為4.9%的模型，並且透過基因演算法後得知當料溫為207.86℃、保壓時間為12.6秒、一段保壓壓力為569.28 bar、二段保壓壓力為569.28 bar、三段保壓壓力為569.28 bar的時候為最佳參數，經過驗證實驗後得到0.02027的反應值，相比之下藉由反應曲面法與田口法優化後之驗證實驗的反應值為0.02032和0.02108。根據以上的研究結果表示，經由神經網路整合基因演算法優化後的製程參數可以降低比容之偏差，說明在本研究中神經網路整合基因演算法的優化能力更勝於反應曲面法及田口法。

This study utilizes the hybrid back propagation neural network (BPNN) and genetic algorithm (GA) to establish a relationship between process parameters and product quality of injection molded product. The main objective is to optimize the parameters in order to minimize the volumetric shrinkage of the product. First, temperature and pressure sensors within the mold are used to monitor the molten state during the experimental process. The collected data is then converted into specific volume values. Additionally, the combination of the index of non-uniform of specific volume and the index of volumetric shrinkage obtained through specific volume are regarded as an overall indicator of volumetric shrinkage and the response value of the design of experiments. After standardization, the back propagation neural network model is trained and employed as the fitness function for the genetic algorithm. Finally, the comparison of optimization is conducted among the hybrid back propagation neural network and genetic algorithm, the response surface method (RSM) as well as Taguchi method.
The results of this study indicate that the neural network model with a 5-11-1 architecture achieves an average absolute percentage error of 4.9% between prediction and measurement on the validation set. After applying the genetic algorithm for optimization, the optimal process parameters in this model are determined as follows: melt temperature of 207.86℃, packing time of 12.6 seconds, the first packing pressure of 569.28 bar, the second packing pressure of 569.28 bar, and the third packing pressure of 569.28 bar. The corresponding response value obtained from the optimal experiment using hybrid ANN and GA is 0.02027, which outperforms the response values of 0.02032 and 0.02108 obtained from the optimal experiments using the response surface methodology and Taguchi method, respectively.
These results demonstrate that the hybrid ANN and GA can reduce the deviation of the specific volume. Furthermore, it indicates the superior optimizing capability of the hybrid ANN and GA compared to response surface method and Taguchi method.

[1] M. Berry and N. Schott, "Process monitoring and process control: an overview," Applied Plastics Engineering Handbook, pp. 377-393, 2017.
[2] B. Pramujati, R. Dubay, and C. Samaan, "Cavity pressure control during cooling in plastic injection molding," Advances in Polymer Technology: Journal of the Polymer Processing Institute, vol. 25, no. 3, pp. 170-181, 2006.
[3] J.-S. Gim, J.-S. Tae, J.-H. Jeon, J.-H. Choi, and B.-O. Rhee, "Detection method of filling imbalance in a multi-cavity mold for small lens," International Journal of Precision Engineering and Manufacturing, vol. 16, pp. 531-535, 2015.
[4] T. Ageyeva, S. Horváth, and J. G. Kovács, "In-mold sensors for injection molding: On the way to industry 4.0," Sensors, vol. 19, no. 16, p. 3551, 2019.
[5] S. Biehl, N. Paetsch, E. Meyer-Kornblum, and G. Brauer, "Wear resistenat thin film sensor system for industrial applications," Int. J. Instrum. Meas, vol. 1, pp. 6-11, 2016.
[6] M. R. Groleau and R. Groleau, "Comparing Cavity Pressure Sensor Technologies Using In-Mold Data (446)," in ANTEC-CONFERENCE PROCEEDINGS-, 2002, vol. 3: UNKNOWN, pp. 3400-3404.
[7] P. Zhao et al., "Intelligent injection molding on sensing, optimization, and control," Advances in Polymer Technology, vol. 2020, pp. 1-22, 2020.
[8] C. Abeykoon, P. J. Martin, A. L. Kelly, and E. C. Brown, "A review and evaluation of melt temperature sensors for polymer extrusion," Sensors and actuators A: Physical, vol. 182, pp. 16-27, 2012.
[9] J.-Y. Chen, K.-J. Yang, and M.-S. Huang, "Online quality monitoring of molten resin in injection molding," International Journal of Heat and Mass Transfer, vol. 122, pp. 681-693, 2018.
[10] S. Farahani, V. Khade, S. Basu, and S. Pilla, "A data-driven predictive maintenance framework for injection molding process," Journal of Manufacturing Processes, vol. 80, pp. 887-897, 2022.
[11] J. Gim and B. Rhee, "Novel analysis methodology of cavity pressure profiles in injection-molding processes using interpretation of machine learning model," Polymers, vol. 13, no. 19, p. 3297, 2021.
[12] W.-C. Chen, G.-L. Fu, P.-H. Tai, and W.-J. Deng, "Process parameter optimization for MIMO plastic injection molding via soft computing," Expert Systems with Applications, vol. 36, no. 2, pp. 1114-1122, 2009.
[13] W.-C. Chen, P.-H. Tai, M.-W. Wang, W.-J. Deng, and C.-T. Chen, "A neural network-based approach for dynamic quality prediction in a plastic injection molding process," Expert systems with Applications, vol. 35, no. 3, pp. 843-849, 2008.
[14] F. Yin, H. Mao, L. Hua, W. Guo, and M. Shu, "Back propagation neural network modeling for warpage prediction and optimization of plastic products during injection molding," Materials & design, vol. 32, no. 4, pp. 1844-1850, 2011.
[15] Y. Cao, X. Fan, Y. Guo, S. Li, and H. Huang, "Multi-objective optimization of injection-molded plastic parts using entropy weight, random forest, and genetic algorithm methods," Journal of Polymer Engineering, vol. 40, no. 4, pp. 360-371, 2020.
[16] X.-P. Li, G.-Q. Zhao, Y.-J. Guan, and M.-X. Ma, "Optimal design of heating channels for rapid heating cycle injection mold based on response surface and genetic algorithm," Materials & Design, vol. 30, no. 10, pp. 4317-4323, 2009.
[17] C. Shen, L. Wang, and Q. Li, "Optimization of injection molding process parameters using combination of artificial neural network and genetic algorithm method," Journal of materials processing technology, vol. 183, no. 2-3, pp. 412-418, 2007.
[18] F. Yin, H. Mao, and L. Hua, "A hybrid of back propagation neural network and genetic algorithm for optimization of injection molding process parameters," Materials & Design, vol. 32, no. 6, pp. 3457-3464, 2011.
[19] R. Chang, C. Chen, and K. Su, "Modifying the tait equation with cooling‐rate effects to predict the pressure–volume–temperature behaviors of amorphous polymers: Modeling and experiments," Polymer Engineering & Science, vol. 36, no. 13, pp. 1789-1795, 1996.
[20] C. Xiao, J. Ye, R. M. Esteves, and C. Rong, "Using Spearman's correlation coefficients for exploratory data analysis on big dataset," Concurrency and Computation: Practice and Experience, vol. 28, no. 14, pp. 3866-3878, 2016.
[21] M. W. Gardner and S. Dorling, "Artificial neural networks (the multilayer perceptron)—a review of applications in the atmospheric sciences," Atmospheric environment, vol. 32, no. 14-15, pp. 2627-2636, 1998.
[22] E. G. Learned-Miller, "Introduction to supervised learning," I: Department of Computer Science, University of Massachusetts, p. 3, 2014.
[23] R. Hecht-Nielsen, "Kolmogorov’s mapping neural network existence theorem," in Proceedings of the international conference on Neural Networks, 1987, vol. 3: IEEE Press New York, NY, USA, pp. 11-14.
[24] J. Wang, "Digital image encryption algorithm design based on genetic hyperchaos," International Journal of Optics, vol. 2016, 2016.
[25] B. Rajakumar and A. George, "APOGA: An adaptive population pool size based genetic algorithm," AASRI Procedia, vol. 4, pp. 288-296, 2013.
[26] A. Géron, Hands-on machine learning with Scikit-Learn, Keras, and TensorFlow. " O'Reilly Media, Inc.", 2022.
[27] H. Motoda and H. Liu, "Feature selection, extraction and construction," Communication of IICM (Institute of Information and Computing Machinery, Taiwan), vol. 5, no. 67-72, p. 2, 2002.
[28] M. Dorigo and L. M. Gambardella, "Ant colony system: a cooperative learning approach to the traveling salesman problem," IEEE Transactions on evolutionary computation, vol. 1, no. 1, pp. 53-66, 1997.
[29] A. Alturki, O. Bchir, and M. M. Ben Ismail, "Depth-Adaptive Deep Neural Network Based on Learning Layer Relevance Weights," Applied Sciences, vol. 13, no. 1, p. 398, 2022.
[30] M. Shafay, R. W. Ahmad, K. Salah, I. Yaqoob, R. Jayaraman, and M. Omar, "Blockchain for deep learning: review and open challenges," Cluster Computing, vol. 26, no. 1, pp. 197-221, 2023.