Reliability-centered prediction methods of clock battery low-output-voltage event for smart electricity meters.

Smart electricity meter failures triggered by low-output-voltage accounted for 79 % of on-site meter failures. It is significant to predict the event of low-output-voltage of the clock battery. Still, it is also challengeable due to the complexity of the failure causes and the diversity of component reliability data. In this paper, multi-causes of battery low-output-voltage failure were analyzed and incorporated to develop a system reliability model for the clock battery circuit. The lifetime distribution of the clock battery circuit was computed using time-varying weights to integrate diverse forms of component data, including degradation data, lifetime distribution data, lifetime observation data, and failure rate data from datasheet. Finally, an engineering example is used to verify the effectiveness of the developed method. The developed reliability model of clock-battery systems can also be applied in other systems where power is supplied from batteries. [ABSTRACT FROM AUTHOR]