Adaptive Neural Network for Autonomous Quality Assurance in Large-Scale Additive Manufacturing: A Comprehensive Approach to Flow Dynamics, Geometric Precision, and Thermal Process

Author(s): Sai Kothapalli

Publication #: 2506025

Date of Publication: 11.02.2025

Country: USA

Pages: 1-10

Published In: Volume 11 Issue 1 February-2025

DOI: https://doi.org/10.5281/zenodo.15699231

Abstract

Construction 3D printing faces significant quality control challenges including incorrect flow rates, layer misalignments, and temperature fluctuations that compromise structural integrity and geometric accuracy. This paper presents a comprehensive machine learning framework integrating computer vision, predictive modeling, and real-time control systems to address these critical issues. This research approach combines convolutional neural networks (CNNs) for visual defect detection, long short-term memory (LSTM) networks for temperature prediction, and reinforcement learning for flow rate optimization. Experimental validation on a large-scale concrete 3D printer demonstrates 87% reduction in flow rate deviations, 92% improvement in layer alignment accuracy, and 78% decrease in temperature fluctuation-induced defects. The proposed system achieves real-time performance with 15 ms response time, enabling immediate corrective actions during the printing process.

Keywords:

Download/View Paper's PDF

Download/View Count: 1

Share this Article