Distributed　robust　consensus　control　problem　is　investigated　for　multi-agent　systems　(MASs)　in　directed　networks　with　dynamics　is　dominanted　by　linear　continuous　models　with　polytopic　uncertainty.　A　linear　transformation　approach　is　applied,　which　transforms　the　state　consensus　problem　of　the　uncertain　MASs　into　the　robust　quadratic　partial　stability　problem　of　a　corresponding　linear　parameter-varying　(LPV)　system.　In　doing　the　linear　transformation,　an　incidence　matrix　is　established　as　the　transformation　matrix,　which　can　make　the　original　uncertain　system　transformed　into　a　reduced-order　system.　According　to　this　decomposed　system,　quadratic　stabilization　system　with　robust　controller　is　designed.　The　controller　construction　mainly　relies　on　solving　a　finite　linear　matrix　inequalities　(LMIs)　to　approximate,　and　in　the　meantime　the　control　gain　matrix　of　the　proposed　distributed　robust　consensus　protocol　is　also　designed.　Finally,　the　effectiveness　of　the　proposed　method　is　verified　by　two　examples.