The　essence　of　the　traditional　multiway　kernel　independent　component　analysis(MKICA)　method　is　that　the　independent　component　analysis(ICA)　whitened　principal　component　analysis(PCA)　is　replaced　with　KPCA　by　using　second　order　statistics　of　the　monitoring　and　controlling　process,　not　by　the　stage　characteristic　of　process　data　and　higher-order　cumulant　information.　To　solve　this　problem,　the　high　order　cumulant　analysis(HCA)　and　multiway　kernel　entropy　independent　component　analysis(MKECA)　are　combined,　and　the　analysis　of　high　order　cumulant　multiway　kernel　entropy　independent　component　analysis(HCA-MKEICA)　method　is　proposed.　Firstyly,　the　kernel　entropy　independent　component　analysis(KECA)　method　is　used　for　original　data　conversion　to　solve　the　problem　of　nonlinear.　Then,　in　the　high-dimensional　kernel　entropy　space,　the　HCA　technology　is　used　to　construct　the　new　statistics　for　process　monitoring.　Finally,　the　proposed　method　is　applied　to　the　microbial　fermentation　process,　and　the　comparison　results　with　the　traditional　methods　show　that　the　proposed　method　can　achieve　a　better　detection,　and　verify　its　effectivess.　©　2017,　Editorial　Office　of　Control　and　Decision.　All　right　reserved.