The embodiment of the invention provides a distribution network
lightning stroke fault detection method and
system. The method comprises the following steps: obtaining
lightning activity data and distribution network fault alarm data; the
lightning activity data and the distribution network fault alarm data are subjected to space-
time correlation analysis, multi-dimensional features of correlation events are extracted, lightning
stroke research and judgment data are generated based on a preset research and judgment model, the lightning
stroke research and judgment data comprise lightning stroke fault judgment results and
tower risk levels, and the multi-dimensional features at least comprise ground lightning density and
terrain factors; based on the
tower risk level and a preset lightning protection strategy
library, generating a lightning protection
treatment strategy for at least one
tower in the associated event; and
processing information including the lightning stroke fault judgment result, the associated tower information and the lightning protection
treatment strategy is sent to an active first-aid repair work
order system or a lightning protection measure treatment work
order system. Through multi-dimensional
feature fusion and modeling study and judgment, the accuracy and reliability of screening real lightning stroke faults are improved, and misjudgment and
delay caused by dependence on artificial experience are reduced.