With　the　continuous　development　of　social　economy,　the　number　of　vehicles　has　increased　year　by　year,　resulting　in　a　series　of　urban　traffic　problems　such　as　traffic　jams　and　frequent　traffic　accidents.　Therefore,　the　prediction　of　traffic　flow　has　gradually　attracted　people＇s　attention.　As　machine　learning　and　deep　learning　technologies　develop　rapidly,　more　and　more　methods　are　being　used　for　traffic　flow　prediction.　However,　these　methods　have　some　shortcomings,　such　as　not　considering　periodicity　and　too　many　parameters.　In　this　paper,　we　propose　the　use　of　the　Gate　Recurrent　Unit　(GRU)　model　to　predict　traffic　flow,　using　the　open-access　data　set　of　England　M20　motorway,　finds　a　GRU　model　that　is　more　consistent　with　traffic　flow　prediction　by　modifying　the　number　of　epochs,　input　sequence　length,　batch　size,　hidden　units　and　learning　rate　through　proper　hyperparameter　tuning,　and　the　accuracy　of　prediction　is　evaluated　by　comparing　RMSE　and　MAE　with　other　models.　The　experimental　results　show　that　the　solution　we　proposed　can　achieve　higher　accuracy　in　predicting　traffic　flow　changes.　©　2023　IEEE.