Volume 10,Issue 01

Strategic Management of Manufacturing Quality: Advanced Detection of Process Anomalies Using Machine Learning Models

Authors

Aynaz Farkhondeh, Jonathan Chen

Abstract
In the strategic management of manufacturing operations, the integration of quality control systems is pivotal for sustaining competitive advantage and operational excellence. Control Chart Patterns (CCPs) are instrumental in this regard, offering a data-driven approach to detect and manage process variability and quality. Recognizing the crucial role of CCPs, this manuscript unveils a cutting-edge machine learning methodology specifically engineered for the strategic oversight of manufacturing quality, with a focus on the accurate identification of various CCPs. By harnessing a blend of smart geometric and statistical feature extraction and a refined neural network model, our proposed method unfolds through a trio of classification stages. Each stage employs radial basis function neural networks (RBFNNs), meticulously calibrated using backpropagation algorithms, to pinpoint a subset of CCPs. The fine-tuning of these networks is achieved via particle swarm optimization, determining the optimal number of radial basis functions and their expansion widths. The core contributions of this research include innovative feature extraction methods, bolstered robustness of RBFNNs, and a comprehensive scope encompassing nine CCPs. This meticulously crafted approach culminates in a resilient and finely tuned analytical engine, adept at navigating the complexities of CCPs. Through simulation, we validate that our approach surpasses existing methods, boasting an exemplary pattern recognition accuracy of 99.5%. This paper represents a significant leap in quality control management, equipping organizations with a robust tool to enhance their manufacturing process integrity.

Keyword: Strategic management, Qquality control, Manufacturing operations, Control chart patterns, Machine learning.

PDF [ 772.98 Kb ] | Endnote File | XML