Chang＇an　Street　is　one　of　the　most　important　arterial　roads　in　Beijing,　and　it　is　one　of　ten　congestion　streets　at　the　same　time.　How　to　solve　this　congestion　problem　become　crucial　after　the　29th　Olympic　Game.　This　paper　proposed　using　fuzzy　coordinated　control　method　to　control　this　two-direction　street,　and　three　signal　control　parameters　consisting　of　the　cycle　length,　green　split　and　offset　were　controlled　systematically　by　three　different　fuzzy　controllers　which　used　queuing　length,　density　and　flow　as　inputs.　In　order　to　overcome　hysteresis　problem　in　fuzzy　method,　we　applied　traffic　flow　forecasting　which　can　forecast　the　coming　flow　in　2min　interval　to　give　flow　input　to　the　fuzzy　controllers.　Clustering　method　was　used　to　organize　the　history　flow　data　which　were　collected　by　the　detectors　located　on　the　arterial　road,　and　then　KNN　method　forecasted　the　coming　flow　in　next　2min　by　using　the　history　data　and　current　detected　data.　Finally,　a　simulation　test　with　a　real-world　data　in　Paramics　software　has　been　done　to　explain　the　advantage　of　the　proposed　method　in　comparison　with　the　fixed-time　method　which　is　being　used　in　Chang＇an　Street.　©　2008　IEEE.