Abstract:

As a circuit protection device, polymer positive temperature coefficient (PPTC) resettable fuses have been widely used in over-current or over-temperature circuit-protection designs in computers, automotive circuits, telecommunications equipment, and medical devices. Failure of a resettable fuse can cause abnormal operation of and damage to a circuit. Prognostics for a resettable fuse enables advance warning of the failure of the fuse and the prediction of its remaining useful life; therefore, the operator can take action to maintain or replace the component to reduce damage to the circuit. This paper uses the cross-validation equential probability ratio test (CV-SPRT) method to monitor the failure precursor parameters in situ, and uses an autoregressive integrated moving average (ARIMA) model to predict failure based on the precursors.