Distribution　system　state　estimation　is　one　of　the　key　elements　for　distribution　network　monitoring　activities.　It　estimates　the　voltage　amplitude　and　phase　angle　of　each　node　on　the　basis　of　redundant　measurement　information.　This　paper　presents　a　dynamic　state　update　method　considering　the　influence　of　the　correlation　among　different　measurement　data.　Firstly,　the　correlation　between　the　amplitude　and　phase　measured　by　the　same　transformer　is　analyzed.　Secondly,　the　covariance　matrix　among　the　measurement　data　can　be　obtained　by　using　unscented　transformation　method,　and　is　applied　to　the　unscented　Kalman　filter　(UKF)　for　calculating　the　current　state.　Finally,　we　assess　the　performance　of　the　proposed　method　on　an　IEEE-33　distribution　network.　The　results　indicate　that　the　method　can　achieve　a　satisfactory　estimation　accuracy　and　avoids　obtaining　suitable　measurement　error　covariance　only　relying　on　experience.　©　2021,　Springer　Nature　Singapore　Pte　Ltd.