Fault Detection on Electronic Boards with MSPRT MSPRT ile Elektronik Kartlarda Hata Tespiti
34th Signal Processing and Communications Applications Conference, SIU 2026, İstanbul, Türkiye, 7 - 10 Temmuz 2026, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/siu71813.2026.11636822
- Basıldığı Şehir: İstanbul
- Basıldığı Ülke: Türkiye
- Anahtar Kelimeler: Fault Detection, Gaussian Distribution, MSPRT, Multi Hypothesis Testing, Real Time Diagnostics, Sequential Analysis, Thermal Imaging, Time Series Analysis
- Marmara Üniversitesi Adresli: Evet
Özet
This study presents a real time fault detection system for circuit boards using thermal imaging sequences. The framework uses time varying Gaussian distributions to model the heating profiles of various components over time. Classification is handled by the Multi Hypothesis Sequential Probability Ratio Test (MSPRT) and it makes diagnostic decisions by collecting statistical data across consecutive frames. To ensure the system is not affected by sensor noise or sudden thermal changes, a stability rule are used. This rule requires the decision threshold to be reached for a specific number of frames in a row N before a final result is confirmed. The method reduces inspection time by identifying the correct class as soon as the evidence meets the threshold.