Quality Prediction in Multi-Stage Manufacturing Systems Based on Deep Neural Networks

Abstract

 In modern manufacturing, quality prediction plays a crucial role in ensuring product reliability and optimizing production efficiency. This study investigates the application of deep neural networks (DNNs) for quality prediction in multi-stage manufacturing processes (MMPs), where complex dependencies exist among stages. By leveraging historical process data and inter-stage feature correlations, we propose an end-to-end DNN-based framework capable of learning nonlinear mappings between process parameters and final product quality. Experimental results on real-world datasets demonstrate that the proposed model outperforms traditional machine learning methods in prediction accuracy and robustness, particularly under conditions with missing or noisy intermediate data. This research provides new insights into inteligent quality control and predictive analytics in smart manufacturing systems.