The　accurate　control　of　automotive　fuel　cell　oxygen　excess　ratio　(OER)　is　necessary　to　improve　system　efficiency　and　service　life.　To　this　end,　an　anti-disturbance　control　driven　by　a　feedback　linearization　model　predictive　control　(MPC)-based　cascade　scheme　is　proposed.　It　considers　strong　nonlinear　coupling　and　disturbance　injection　of　fuel　cell　oxygen　supply.　A　six-order　nonlinear　fuel　cell　oxygen　feeding　model　is　presented.　It　is　further　formulated　using　an　extended　state　observer　to　rapidly　reconstruct　the　OER,　to　overcome　the　slow　response　and　interference　errors　of　sensor　measurements.　In　the　proposed　cascade　control,　the　outer　loop　is　the　anti-disturbance　control　which　is　used　to　realize　the　optimized　OER　tracking　and　the　inner　loop　via　the　feedback　linearization　to　linearize　the　oxygen　feeding　behaviors　conducts　MPC　to　regulate　the　air　compressor　output　mass　flow.　The　feedback　linearization　demonstrates　a　robust　tracking　performance　of　nonlinear　outputs,　and　the　integral　absolute　error　of　anti-disturbance　control　is　0.3021　lower　than　that　of　PI　control　under　a　custom　test　condition.　Finally,　the　numerical　validation　on　a　hybrid　driving　cycle　indicates　that　the　proposed　cascade　control　can　regulate　the　fuel　cell　OER　with　an　average　absolute　error　of　0.02313　in　the　high　air　compressor　operation　efficiency　zone.　(c)　2020　Hydrogen　Energy　Publications　LLC.　Published　by　Elsevier　Ltd.　All　rights　reserved.