Membrane　fouling　is　a　widespread　problem　that　restricts　the　stable　operation　of　membrane　bioreactor　(MBR)　in　wastewater　treatment　process　(WWTP).　However,　it　is　difficult　to　avoid　the　occurrence　of　membrane　fouling　due　to　the　lack　of　effective　early　warning　methods.　To　deal　with　this　problem,　an　intelligent　early　warning　method,　using　a　knowledge-based　fuzzy　broad　learning　(K-FBL)　algorithm,　is　proposed　for　membrane　fouling　in　this　paper.　First,　the　existing　knowledge　is　extracted　from　the　humanistic　category　of　membrane　fouling　in　the　form　of　fuzzy　rules.　Then,　the　existing　knowledge　of　membrane　fouling　can　be　used　to　compensate　for　the　shortage　of　data　sets.　Second,　a　K-FBL　algorithm　is　designed　to　train　the　fuzzy　subsystems　with　the　existing　knowledge.　Then,　the　uncertainties　of　membrane　fouling　process　can　be　degraded　to　improve　the　learning　performance.　Third,　a　K-FBL-based　early　warning　method　is　designed　to　realize　the　precise　classification　and　provide　the　operational　suggestions　for　membrane　fouling.　Finally,　the　experiment　results　of　a　real　plant　are　given　to　demonstrate　the　effectiveness　of　this　proposed　K-FBL-based　early　warning　method.