Current　research　on　knowledge　graphs　focuses　mostly　on　static　knowledge　graphs　while　ignoring　temporal　information.　Recently,　people　have　begun　to　study　the　temporal　knowledge　graph,　which　integrates　temporal　information　into　KGC,　so　that　the　modeling　is　constantly　changing　with　the　knowledge　that　evolves　over　time.　In　this　survey,　we　summarize　the　existing　temporal　knowledge　graph　research,　which　is　divided　into　extrapolation　tasks　and　interpolation　tasks　according　to　time.　The　extrapolation　task　is　mainly　used　to　predict　future　facts　and　consists　of　three　models:　Temporal　Point　Process,　Time　Series,　and　other　models.　The　interpolation　task　extends　the　existing　KGC　models　to　complement　the　lack　of　past　temporal　information,　including　five　models:　Translational　Distance,　Semantic　Matching,　Neural,　Relational　Rotation,　and　Hyperbolic　Geometric　models.