Methods　based　on　multivariate　statistical　projection　analysis　have　been　widely　applied　for　batch　processes　monitoring.　However,　conventional　methods　are　linear　ones　that　can　only　model　linear　combinations　of　variables　and　most　batch　processes　are　non-linearity.　Traditionally,　in　process　modeling,　two　solutions　for　non-linearity　have　been　implemented:　non-linear　models　and　local　linear　models.　In　this　paper,　a　novel　methodology　named　Sub-phase　based　Principal　Component　Analysis　(SPPCA),　which　integrates　methods　of　operation　phase　detection　and　a　novel　multi-way　principal　component　analysis　(MPCA),　is　approached.　A　case　study　from　a　simulated　fed-batch　penicillin　cultivation　process　indicates　the　efficacy　of　approach.　©　2014　TCCT,　CAA.