The　model　establishment　is　of　great　significance　to　realize　the　automatic　operation　of　membrane　bioreactor　(MBR).　In　this　study,　a　multi-functional　rinsing　model　based　on　the　variation　of　cake　properties　was　established　to　predict　the　instantaneous　fouling　resistance/rinsing　efficiency　(η)　of　0.1　μm　PAN　membranes　fouled　with　activated　sludge　suspension　and　calculate　the　energy　consumption　(W)　for　cake　removal.　The　results　showed　that　the　model　predictions　had　good　agreements　with　experimental　data　(R2＞0.960).　Meanwhile,　the　cake　thickness　(L)/porosity　(Εc)　decreased　while　the　specific　cake　resistance　(r)　increased　in　the　ringing　process.　And　the　cake　with　smaller　Εc　and　larger　r　was　more　difficult　to　remove.　In　addition,　the　transition　points　of　W　(uw)　and　η　(uη)　were　determined　(uw　η).　And　the　variation　trend　of　η　and　W　could　be　divided　into　three　stages:　(i)　η　increased　rapidly　while　W　increased　slowly　before　uw;　(ii)　η　increased　slowly　while　W　increased　rapidly　in　the　range　from　uw　to　uη;　(iii)　η　increased　slightly　while　W　increased　more　rapidly　after　uη.　The　suitable　cross-flow　velocity　range　(uw　η)　and　the　suitable　working　region　were　determined　after　weighing　W　and　η,　which　can　be　used　to　select　optimal　rinsing　conditions　in　the　actual　MBR.　©　2022　Elsevier　B.V.