Reliability Prediction of Thyristor using Artificial Intelligence Techniques

Objective: To use temperature and duty cycle as a health assessment tool for implementing prognostics and health management in thyristor and which can be further implemented to various other electronics components. Methods: To predict the reliability of thyristor, statistical approach has been identified and analysed for its life prediction. The military handbook MIL-HDBK-217F N2 and Reliability Information Analysis centre RIAC 217Plus™ failure models are explored. Further in order to verify and validate the effectiveness and reliability of these systems, both MIL-HDBK-217F N2 and RIAC 217Plus™ reliability results have been compared. Findings: The operational parameters and modelling of thyristor have been selected and life prediction analysis is carried out. The analysis has been identified by using statistical methods. The various conditions such as temperature, stress and design parameters have been explored and screening has been done to improve reliability. It has been analysed that input current and temperature are the decisive parameters that affects the health and failure mechanism of thyristor. The study of model parameter with respect to variation in stress parameters have been carried out. Novelty: To validate the effectiveness of developed system, both MIL-HDBK-217F N2 and RIAC 217Plus™ assessment prediction equations are compared.