Belief　plays　an　imortant　part　in　intelligent　systems　modeling.　On　the　one　hand,　in　the　multi-agent　systems　community,　belief　is　usually　modeled　as　a　mental　state　of　agents　by　modal　logic　in　the　context　of　BDI　(beliefdesire　-　intention).　On　the　other　hand,　in　epistemic　logic　research,　belief　and　knowledge　are　often　taken　into　account　together,　which　are　main　components　of　knowledge-based　systems.　However　problems　such　as　logical　omniscience　and　several　others　come　along　with　this　one　still　left　unsettled　since　agents　are　resource　bounded　reasoners.　In　this　paper,　the　construction　of　fuzzy　belief　logic　based　on　plausibility　degree　is　considered,　three　kinds　of　fuzzy　belief　operators　(individual　fuzzy　belief,　group　fuzzy　belief,　and　common　fuzzy　belief)　are　investigated,　and　the　associated　Kripke　semantics　is　given.　Logical　omniscience　problem　of　belief　modeling　is　done　away　with　and　proved.　So　that　a　new　powerful　tool　for　intelligent　systems　modeling　is　well　constructured.　©　2008　IEEE.