To address the limitations of traditional quality management methods in processing and analyzing massive data for electronic components，this study aims to establish an intelligent quality monitoring mechanism for enhancing the accuracy and reliability of quality assessment. A novel dual-model framework integrating quality monitoring and fault prediction was established using the whole-life-cycle multi-source data: 1）A quality assessment model employing hyperparameter-optimized machine learning algorithms was constructed，utilizing six-dimensional feature data covering factory inspection，in-process quality assurance，and defect records；2）A fault prediction model was designed based on a backpropagation（BP）neural network to enable dynamic early warnings. Experimental validation on JZC-084 electromagnetic relays and J599F26D low-frequency connectors demonstrated that the proposed method achieved a fault prediction error rate lower than 0.1% and a quality assessment accuracy of 95.1%，which exceeded technical specifications. Verification via the random forest classifier showed average precision，recall，and F1-score values of 83.6%，81.2%，and 78.3%，respectively. This data-driven approach significantly enhances scientific decision-making in quality management through real-time monitoring and cross-departmental data synergy. Future work will focus on model parameter optimization and scenario expansion to enhance prediction comprehensiveness.