Adaptive quality prediction in injection molding based on ensemble learning

Abstract Despite the central role of quality, industry applications of data-based quality prediction for thermoplastics injection molding are rare, because of a suboptimal cost-benefit ratio. Therefore, we present a holistic approach for seamless part quality prediction, which automates the necessary data processing steps. Since the performance of the seven supervised learning algorithms applied with Bayesian hyperparameter-optimization depends on aspects such as process state, etc., we combine the learnt models using an ensemble-method to ensure good results under varying conditions. The results show that the ensemble’s performance significantly depends on the chosen ensemble-hyperparameters, so future research should focus on their automatic identification.

[1]  Gaston H. Gonnet,et al.  Automatic problem-specific hyperparameter optimization and model selection for supervised machine learning , 2015 .

[2]  Sylvain Arlot,et al.  A survey of cross-validation procedures for model selection , 2009, 0907.4728.

[3]  Vladimir Vapnik,et al.  Support-vector networks , 2004, Machine Learning.

[4]  Tobias Meisen,et al.  Transfer-Learning: Bridging the Gap between Real and Simulation Data for Machine Learning in Injection Molding , 2018 .

[5]  Olga Ogorodnyk,et al.  Application of Machine Learning Methods for Prediction of Parts Quality in Thermoplastics Injection Molding , 2018 .

[6]  A. G. Asuero,et al.  The Correlation Coefficient: An Overview , 2006 .

[7]  Klaus-Dieter Thoben,et al.  Machine learning in manufacturing: advantages, challenges, and applications , 2016 .

[8]  Hugues Favrelière,et al.  Generative Adversarial Networks for geometric surfaces prediction in injection molding , 2018, ICIT 2018.

[9]  Maurice Pillet,et al.  Quality prediction in injection molding , 2017, 2017 IEEE International Conference on Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA).

[10]  Reinhard Schiffers.  Verbesserung der Prozessfähigkeit beim Spritzgießen durch Nutzung von Prozessdaten und eine neuartige Schneckenhubführung , 2009 .

[11]  António Gaspar-Cunha,et al.  Modeling and optimization of the injection-molding process: a review , 2018 .

[12]  Olga Ogorodnyk,et al.  Application of feature selection methods for defining critical parameters in thermoplastics injection molding , 2019 .

[13]  Bernhard Schölkopf,et al.  A tutorial on support vector regression , 2004, Stat. Comput..

[14]  Reinhard Schiffers,et al.  A Holistic Approach to Part Quality Prediction in Injection Molding Based on Machine Learning , 2020 .

[15]  Tony Wang,et al.  Product Attribute Forecast: Adaptive Model Selection Using Real-Time Machine Learning , 2018 .

[16]  Ferat Sahin,et al.  A survey on feature selection methods , 2014, Comput. Electr. Eng..

[17]  Nikola Bogunovic,et al.  A review of feature selection methods with applications , 2015, 2015 38th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO).

[18]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[19]  Reinhard Schiffers,et al.  Comparison of feature selection methods for machine learning based injection molding quality prediction , 2020 .

[20]  Mark A. Hall,et al.  Correlation-based Feature Selection for Machine Learning , 2003 .

[21]  Adam Krzyzak,et al.  An affine invariant k-nearest neighbor regression estimate , 2012, J. Multivar. Anal..