Quality Monitoring of Production Process with Unknown Parameter Distribution

The development of modern industry demands that factories, amidst a flexible production model involving mixed large and small batch production of multiple varieties, research precise real-time quality monitoring methods to face the risks of quality fluctuations due to frequent process adjustments, ensuring consistent product quality.This study aimed to integrate statistical process control (SPC) for monitoring quality anomalies in processes with unknown parameter distributions. We developed a non-parametric GWMA SR control chart for real-time quality monitoring. For anomaly recognition, we performed feature selection and fusion, reduced feature dimensionality with PCA, and employed GA-optimized RF for pattern classification. Simulation tests compared recognition speed and accuracy across feature combinations and classification methods, demonstrating the proposed method's superior performance. Applied to UO₂ pellet diameter monitoring, the combined GWMA SR control chart and PCA-RF-GA model effectively identified abnormal quality fluctuations in actual production, confirming its accuracy and speed in managing processes with unknown parameter distributions.