With　the　proposal　of　the　theory　of　“smart　city”,　the　public　transportation　travel　information　service　is　facing　the　requirement　of　higher　quality,　and　the　optimization　of　the　traffic　data　used　to　support　the　travel　information　service　in　the　process　of　collection　and　processing　is　the　key　and　difficult　point.　In　view　of　the　current　data　source　used　to　support　the　public　transportation　information　service　calculation　to　the　calculation　only　based　on　bus　GPS　data,　and　the　data　source　data　less,　long　sampling　interval,　road　calculation　accuracy　is　low,　combined　with　the　current　traffic　information　fill　and　data　fusion　technology　research　and　development,　in　the　analysis　of　bus　data　and　taxi　data　commonness　and　characteristics,　the　rule　of　historical　data　and　the　connection　between　real-time　data,　puts　forward　a　set　of　based　on　traffic　information　to　fill　and　data　fusion　algorithm　for　bus　arrival　time　provide　data　support　prediction.　Through　the　actual　operation　data　test,　the　data　fusion　algorithm　proposed　in　this　paper　reduced　the　average　relative　error　of　bus　travel　time　by　8%,　introduced　the　traffic　information　filling　model　and　data　fusion　model　into　the　specific　bus　arrival　system　for　actual　test,　and　the　final　prediction　accuracy　reached　86.27　and　7.　09%.　The　above　experimental　data　show　that　the　proposed　method　has　significant　advantages　and　good　applicability　over　the　existing　methods.　©　The　Author(s),　under　exclusive　license　to　Springer　Nature　Singapore　Pte　Ltd.　2024.