Fault diagnosis using cross-wavelet transform for electricity pre-payment meter

Abstract

Direct identification and subsequent replacement of faulty components in electronic circuits reduces the overall maintenance costs and time-to-repair the items. In this paper, a tool that can identify the faulty component in any position in the circuit has been designed, simulated and implemented. Device software use wavelet transform algorithm to recognize the signature of the faulty component. The circuit is broken down to its sub-circuits that each one has a stimuli point and tests nodes, so that each component within the sub-circuit response by a distinguishable signature among other component of the same sub-circuit when it fails. The system was tested and a high level of fault detection was recorded.