Lithium thionyl chloride (Li/SOCl2) batteries are the core backup power source for smart energy meters

and their operational reliability is the guarantee of accurate metering for tariff calculation and time-stepped power metering

but due to the complex structural characteristics of Li/SOCl2 batteries and the influence of hardware environmental factors

it is a challenge to realize accurate prediction of their reliability.In this paper

we review the methods of Li/SOCl2 battery reliability prediction

systematically analyze the existing prediction methods and technologies from three aspects:traditional experimental methods

model-based prediction methods and data-driven prediction methods

summarize the advantages of various battery residual life prediction methods

the remaining problems

and the need to integrate the advantages of multi-methods in the future to optimize experimental techniques

enhance the coupling of the model

and improve the interpretability of the data-driven model

in order to ensure the accuracy of time-sharing ladder metering.In the future

it is necessary to integrate the advantages of multiple methods

optimize the experimental technology

enhance the model coupling

and improve the interpretability of the data-driven model

so as to improve the prediction accuracy

and then provide theoretical support for the long-term and stable operation of smart meters.