With　the　gradual　increase　of　renewable　energy　penetration,　the　operation　situation　of　the　integrated　electricity　and　gas　system　becomes　more　complex　and　variable.　An　affine　arithmetic-based　optimal　power-gas　flow　online　calculation　approach　is　proposed　to　respond　to　the　real-time　changes　of　source　and　load.　Firstly,　a　natural　gas　dynamic　model　is　constructed　using　the　Taylor　expansion　and　implicit　finite　difference　method,　and　the　model＇s　accuracy　is　improved　by　using　a　flow　velocity　base　value　adaptive　correction　method.　Then,　affine　arithmetic　is　utilized　to　describe　the　source-load　uncertainty　of　the　system,　the　optimal　energy　flow　is　expanded　from　the　deterministic　domain　to　the　affine　domain,　and　the　affine　arithmetic-based　optimal　power-gas　flow　model　is　established.　On　this　basis,　the　affine　analytical　expression　of　the　optimal　dispatch　scheme　concerning　the　noise　symbol　is　obtained　through　offline　optimization,　and　the　real-time　change　of　source　and　load　using　the　noise　symbol　is　tracked　to　obtain　the　decision-making　scheme　online.　The　proposed　method　can　meet　the　needs　of　real-time　dispatching　by　making　quick　decisions　based　on　immediate　information.　Finally,　the　case　studies　verify　the　accuracy　of　the　proposed　dynamic　model　and　show　that　the　dispatching　performance　of　the　proposed　online　method　is　close　to　the　optimal　dispatching　with　complete　prediction　information.　©　2024　Power　System　Technology　Press.　All　rights　reserved.