Freeway　traffic　state　estimation　is　the　major　assurance　of　traffic　management　and　control.　This　paper　addresses　a　traffic　density　estimator　for　Beijing　third　ring　freeway.　First,　we　apply　a　new　modeling　method　called　Dynamic　Graph　Hybrid　Automata　(DGHA)　to　express　the　traffic　dynamics　and　the　evolution　law　of　the　freeway　network　by　piecewise-linear.　On　this　basis,　it　is　feasible　to　use　a　open　source　software　named　Open-Modelica　to　establish　Beijing　third　ring　freeway　network　model.　However,　the　presence　of　noise　in　detected　data　from　sensors　increases　the　difficulty　of　density　estimation,　so　a　Kalman　filter　is　employed　to　obtain　the　high　accuracy　density　of　each　network　section　based　on　the　freeway　network　model.　Finally　the　proposed　estimate　solution　is　validated　through　simulation　results,　illustrating　the　ability　and　potential　of　the　density　estimation.　©　2016　IEEE.