In　this　paper,　we　introduce　a　new　hybrid　fuzzy　clustering　approach　for　Takagi-Sugeno　(TS)　fuzzy　modeling　based　on　a　hybrid　fuzzy　clustering　scheme.　The　method　consists　of　a　sequence　of　the　steps　aiming　towards　construct　an　optimal　Takagi-Sugeno　fuzzy　model　from　sample　data.　First,　we　apply　the　modified　mountain　clustering　algorithm　to　automatically　determine　the　number　of　cluster　and　initial　cluster　center.　Second,　the　initial　input-output　space　fuzzy　partition　matrix　is　constructed　by　the　initial　cluster　centers　and　then　the　improved　Gustafson-Kessel　clustering　algorithm　is　utilized　to　result　in　an　optimal　input-output　space　fuzzy　partition　matrix.　Finally,　the　Particle　Swarm　Optimization　(PSO)　algorithm　is　utilized　to　fine　tune　the　system　parameters.　Compared　with　other　fuzzy　modeling　methods,　the　introduced　method　has　the　advantages　of　simplicity,　high　accuracy　and　can　be　handled　by　an　automatic　procedure.　Numerical　examples　are　provided　to　illustrate　the　performance　of　the　proposed　approach.