A　major　issue　in　traffic　control　systems　is　the　high　level　of　uncertainty　due　to　traffic　volume　variation　in　the　dynamics　of　vehicular　queue　and　signal　timings.　An　approach　is　proposed　to　deal　with　the　problem　based　on　the　supervised　multi-model　signal　adaptive　predictive　control　(SMM-SAPC).　According　to　the　characteristics　in　traffic　flow,　such　as　nominal　period,　peak　period　and　a　super　flow　period,　a　supervised　multi-model　approach　for　modeling　the　dynamic　traffic　flow　is　proposed.　By　incorporating　the　traffic　modeling　method　within　MPC　with　control　traffic　signal　systems,　a　novel　intelligent　traffic　control　is　implemented.　Corresponding　response　will　be　made　for　different　traffic　conditions;　an　adaptive　signal　control　for　intersection　in　a　main　road　can　be　implemented.　The　presented　simulations　are　indicative　for　the　reasonable　traffic　time　and　reduction　in　delay　time　and　stop　time　that　can　be　achieved　by　the　proposed　method.