A　general　framework　for　integrating　extension　and　multi-level　knowledge　modeling　is　constructed　to　address　the　problems　existing　in　the　process　of　bionic　prototype　acquisition.　The　multi-level　semantic　relationship　of　the　product-biological　system　is　introduced.　The　extensional　model　of　product　system　Function-Principle-Behavior-Structure　and　biology　system　Structure-　Action-Tactic-Function　case-element　ontology　based　on　the　basic-element　theory　of　the　extension　and　its　formal　description　method　are　built.　The　product　FPBS　and　the　biology　SATF　mapping　relation　model　are　established.　The　case-element　extension　set　of　the　product-biology　system　is　made　according　to　the　extended　nature　and　the　similar　method　of　the　basic-element　is　introduced　in　the　relevant　extensional　set.　Searching　a　bionic　prototype　based　on　the　similarity　of　the　two　systems　is　proposed,　and　a　method　of　extensional　comprehensive　evaluation　is　used　to　optimize　the　bionic　prototype.　The　screen　filter　is　taken　as　an　example,　the　feasibility　of　the　overall　framework　is　verified.　©　2019　Journal　of　Mechanical　Engineering.